CN109255070A - Recommendation information processing method, device, computer equipment and storage medium - Google Patents
Recommendation information processing method, device, computer equipment and storage medium Download PDFInfo
- Publication number
- CN109255070A CN109255070A CN201810865257.1A CN201810865257A CN109255070A CN 109255070 A CN109255070 A CN 109255070A CN 201810865257 A CN201810865257 A CN 201810865257A CN 109255070 A CN109255070 A CN 109255070A
- Authority
- CN
- China
- Prior art keywords
- recommendation information
- vector
- recommendation
- determining
- information
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
- 230000010365 information processing Effects 0.000 title claims abstract description 57
- 238000003672 processing method Methods 0.000 title claims abstract description 41
- 239000013598 vector Substances 0.000 claims abstract description 290
- 238000000034 method Methods 0.000 claims abstract description 26
- 238000004590 computer program Methods 0.000 claims description 9
- 239000012141 concentrate Substances 0.000 abstract 1
- 238000010586 diagram Methods 0.000 description 10
- 230000006870 function Effects 0.000 description 8
- 238000012545 processing Methods 0.000 description 7
- 230000003287 optical effect Effects 0.000 description 6
- 238000013507 mapping Methods 0.000 description 4
- 230000008569 process Effects 0.000 description 4
- 238000012935 Averaging Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 239000000463 material Substances 0.000 description 2
- 239000013307 optical fiber Substances 0.000 description 2
- 230000002093 peripheral effect Effects 0.000 description 2
- 230000000644 propagated effect Effects 0.000 description 2
- 241001481833 Coryphaena hippurus Species 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000003491 array Methods 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 150000007942 carboxylates Chemical class 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 230000002349 favourable effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003062 neural network model Methods 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V10/00—Arrangements for image or video recognition or understanding
- G06V10/40—Extraction of image or video features
- G06V10/46—Descriptors for shape, contour or point-related descriptors, e.g. scale invariant feature transform [SIFT] or bags of words [BoW]; Salient regional features
- G06V10/462—Salient features, e.g. scale invariant feature transforms [SIFT]
- G06V10/464—Salient features, e.g. scale invariant feature transforms [SIFT] using a plurality of salient features, e.g. bag-of-words [BoW] representations
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Computer Vision & Pattern Recognition (AREA)
- General Physics & Mathematics (AREA)
- Data Mining & Analysis (AREA)
- Artificial Intelligence (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Life Sciences & Earth Sciences (AREA)
- General Engineering & Computer Science (AREA)
- Multimedia (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The application proposes a kind of recommendation information processing method, device, computer equipment and storage medium, wherein method comprises determining that the corresponding primary vector of i-th of recommendation information in recommendation list, wherein i is positive integer；According to the primary vector, the corresponding secondary vector of i recommendation information before determining；According to the primary vector and the secondary vector, determine that Candidate Recommendation information concentrates the click probability of each Candidate Recommendation information；According to the click probability of each Candidate Recommendation information, is concentrated from the Candidate Recommendation information and choose i+1 recommendation information.Pass through this method, it realizes according to the relevance between recommendation information and recommendation information is ranked up, be conducive to improve the clicking rate of recommendation information in recommendation list, it solves and recommendation list is determined according to single marking result in the prior art, have ignored the low technical problem of relevance, the clicking rate of recommendation information between recommendation information.
Description
Technical Field
The present application relates to the field of internet technologies, and in particular, to a method and an apparatus for processing recommended information, a computer device, and a storage medium.
Background
With the development of intelligent terminals and internet technologies, information recommendation has become one of the main ways for users to acquire information.
At present, when an information recommendation system carries out information recommendation, a scoring model is mostly used for scoring a plurality of information to be recommended to a user, then the information to be recommended is sorted according to a scoring result, a recommendation list is formed by selecting a preset number of information to be sorted in front, and the recommendation list is fed back to the user.
However, in the information recommendation method, only optimization of a single piece of recommendation information is considered, and relevance between pieces of recommendation information in the recommendation list is ignored, so that the click rate of each piece of recommendation information in the recommendation list is low.
Disclosure of Invention
The present application is directed to solving, at least to some extent, one of the technical problems in the related art.
Therefore, the application provides a recommendation information processing method, a recommendation information processing device, a computer device and a storage medium, which are used for solving the technical problems that in the prior art, a recommendation list is determined according to a single scoring result, the relevance among recommendation information is ignored, and the click rate of the recommendation information is low.
In order to achieve the above object, an embodiment of a first aspect of the present application provides a recommendation information processing method, including:
determining a first vector corresponding to ith recommendation information in a recommendation list, wherein i is a positive integer;
determining second vectors corresponding to the first i pieces of recommendation information according to the first vectors;
determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the first vector and the second vector;
and selecting the (i + 1) th recommendation information from the candidate recommendation information set according to the click probability of each candidate recommendation information.
According to the recommendation information processing method, the first vector corresponding to the ith recommendation information in the recommendation list is determined, the second vector corresponding to the first i recommendation information is determined according to the first vector, the click probability of each candidate recommendation information in the candidate recommendation information set is further determined according to the first vector and the second vector, and the (i + 1) th recommendation information is selected from the candidate recommendation information set according to the click probability of each candidate recommendation information. Therefore, the current recommendation information is determined according to the vector of the previous recommendation information and the vectors of all the recommendation information which are sequenced, the recommendation information is sequenced according to the relevance between the recommendation information, the click rate of the recommendation information in the recommendation list is improved, and the user experience is improved.
To achieve the above object, a second embodiment of the present application proposes a recommended information processing apparatus, including:
the device comprises a first determining module, a second determining module and a third determining module, wherein the first determining module is used for determining a first vector corresponding to the ith recommendation information in a recommendation list, and i is a positive integer;
the second determining module is used for determining second vectors corresponding to the first i pieces of recommendation information according to the first vectors;
the probability determining module is used for determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the first vector and the second vector;
and the selection module is used for selecting the (i + 1) th recommendation information from the candidate recommendation information set according to the click probability of each candidate recommendation information.
According to the recommendation information processing device, the first vector corresponding to the ith recommendation information in the recommendation list is determined, the second vector corresponding to the previous i recommendation information is determined according to the first vector, the click probability of each candidate recommendation information in the candidate recommendation information set is further determined according to the first vector and the second vector, and the (i + 1) th recommendation information is selected from the candidate recommendation information set according to the click probability of each candidate recommendation information. Therefore, the current recommendation information is determined according to the vector of the previous recommendation information and the vectors of all the recommendation information which are sequenced, the recommendation information is sequenced according to the relevance between the recommendation information, the click rate of the recommendation information in the recommendation list is improved, and the user experience is improved.
To achieve the above object, a third aspect of the present application provides a computer device, including: a processor and a memory; wherein the processor executes a program corresponding to the executable program code by reading the executable program code stored in the memory, so as to implement the recommended information processing method according to the embodiment of the first aspect.
To achieve the above object, a fourth aspect of the present application provides a non-transitory computer-readable storage medium, on which a computer program is stored, where the computer program, when executed by a processor, implements the recommended information processing method according to the first aspect.
To achieve the above object, a fifth aspect of the present application provides a computer program product, where instructions of the computer program product, when executed by a processor, implement the recommended information processing method according to the first aspect.
Additional aspects and advantages of the present application will be set forth in part in the description which follows and, in part, will be obvious from the description, or may be learned by practice of the present application.
Drawings
The foregoing and/or additional aspects and advantages of the present application will become apparent and readily appreciated from the following description of the embodiments, taken in conjunction with the accompanying drawings of which:
fig. 1 is a schematic flowchart of a recommended information processing method according to an embodiment of the present application;
FIG. 2 is an exemplary diagram of a position relationship between recommendation information in a recommendation list;
fig. 3 is a schematic flowchart of another recommended information processing method according to an embodiment of the present application;
fig. 4 is a schematic flowchart of another recommended information processing method according to an embodiment of the present application;
fig. 5 is a schematic flowchart of another recommended information processing method according to an embodiment of the present application;
FIG. 6 is a diagram illustrating an exemplary architecture of a recommended information ranking model for implementing the recommended information processing method of the present application;
fig. 7 is a schematic structural diagram of a recommended information processing apparatus according to an embodiment of the present application;
fig. 8 is a schematic structural diagram of another recommended information processing apparatus according to an embodiment of the present application;
fig. 9 is a schematic structural diagram of another recommended information processing apparatus according to an embodiment of the present application;
fig. 10 is a schematic structural diagram of another recommended information processing apparatus according to an embodiment of the present application; and
fig. 11 is a schematic structural diagram of a computer device according to an embodiment of the present application.
Detailed Description
Reference will now be made in detail to embodiments of the present application, examples of which are illustrated in the accompanying drawings, wherein like or similar reference numerals refer to the same or similar elements or elements having the same or similar function throughout. The embodiments described below with reference to the drawings are exemplary and intended to be used for explaining the present application and should not be construed as limiting the present application.
A recommended information processing method, an apparatus, a computer device, and a storage medium according to embodiments of the present application are described below with reference to the drawings.
Fig. 1 is a flowchart illustrating a recommended information processing method according to an embodiment of the present application.
As shown in fig. 1, the recommended information processing method may include the steps of:
step 101, determining a first vector corresponding to the ith recommendation information in the recommendation list, wherein i is a positive integer.
The recommendation list is used for displaying the recommendation information which is sequenced, and when the recommendation information is predicted, the recommendation information is displayed in the recommendation list every time one recommendation information is predicted; the ith recommendation information is the latest recommendation information shown in the recommendation list.
In this embodiment, when predicting recommendation information, a first vector corresponding to the ith recommendation information in the recommendation list may be determined first.
As a possible implementation manner, for text information, when determining a corresponding vector, statistics may be performed on words included in the candidate recommendation information, and a word bag list is constructed according to nonrepeating words included in the candidate recommendation information.
For example, assume that the candidate recommendation information is as follows:
text 1: my dog ate My homework.
Text 2: my carboxylate the sandwich.
Text 3: a dolphin ate the homework.
Then, according to the candidate recommendation information, a bag-of-words list can be constructed as follows: [ a ate a < n > c.
Furthermore, a first vector corresponding to the ith recommendation information can be determined according to the ith recommendation information and the constructed bag-of-words list based on one-hot coding. And the dimension of the first vector is consistent with the length of the bag-of-words list.
Still by taking the above example as an example, assuming that the ith recommendation information is text 1, the first vector corresponding to text 1 is [ 010011100 ].
As a possible implementation manner, the ith recommendation information may also be converted into a corresponding vector by using an existing model, for example, word2vec, so as to obtain a first vector.
It should be noted here that, in addition to the text information, some recommendation information preferably includes picture information, and for the contained picture information, picture features may be extracted, and a corresponding picture vector is determined according to the extracted picture features, so that a first vector corresponding to the recommendation information is obtained according to the text vector corresponding to the text information and the picture vector corresponding to the picture information. For example, the text vector and the picture vector may be added by a summation method to obtain the first vector.
And step 102, determining a second vector corresponding to the previous i pieces of recommendation information according to the first vector.
In this embodiment, after the first vector of the ith recommendation information is determined, a second vector corresponding to the first i recommendation information in the recommendation list may be further determined according to the first vector.
As a possible implementation manner, when the second vector is determined, a weight corresponding to each piece of recommendation information may be determined according to a position relationship between each piece of recommendation information in the recommendation list and the (i + 1) th piece of recommendation information in the recommendation list. For example, according to the presentation style of the recommendation list, the position of the (i + 1) th recommendation information in the recommendation list is determined, and then the weight corresponding to the i previous recommendation information is determined according to the position relationship between the i previous recommendation information and the (i + 1) th recommendation information. Wherein, the closer the distance to the (i + 1) th recommendation information is, the larger the corresponding weight value is.
For example, assuming that the capacity of the recommendation list is 8, 8 pieces of recommendation information can be shown, and the position relationship of the 8 pieces of recommendation information in the recommendation list is shown in fig. 2. Assuming that i is 3, that is, the 4 th recommendation information to be predicted currently, as can be seen from fig. 2, the 2 nd recommendation information and the 3 rd recommendation information are adjacent to the region showing the 4 th recommendation information, and the distance between the 1 st recommendation information and the region showing the 4 th recommendation information is longer, a smaller weight may be allocated to the 1 st recommendation information, and a larger weight may be allocated to the 2 nd recommendation information and the 3 rd recommendation information, for example, the weight corresponding to the 1 st recommendation information is 0.2, and the weights corresponding to the 2 nd recommendation information and the 3 rd recommendation information are both 0.4.
And then, according to the weight and the vector corresponding to each piece of recommendation information, determining a second vector corresponding to the first i pieces of recommendation information. The vector corresponding to each piece of recommendation information in the recommendation list can be determined in the same manner as the first vector corresponding to the ith piece of recommendation information, and for the ith piece of recommendation information, the corresponding vector is the first vector. For the first i pieces of recommendation information, the product of the vector corresponding to each piece of recommendation information and the weight may be calculated first, and then the obtained i results are added to obtain a second vector.
As a possible implementation manner, when determining the weight value corresponding to each piece of recommendation information in the first i pieces of recommendation information, different weight values may be allocated to each piece of recommendation information according to the sequence of each piece of recommendation information. For example, the earlier the order of the recommendation information is, the smaller the corresponding weight. For example, assuming that i is 3, the corresponding weight value may be 0.2 for the 1 st recommendation information, 0.3 for the 2 nd recommendation information, and 0.5 for the 3 rd recommendation information. And then, according to the weight value and the vector corresponding to each piece of recommendation information, calculating to obtain a second vector in a weighted summation mode.
And 103, determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the first vector and the second vector.
The candidate recommendation information set can be determined according to search terms input by a user, and search results recalled according to the search terms are used as the candidate recommendation information set; alternatively, the candidate recommendation information set may be configured by determining based on a historical browsing record of the user, for example, determining a plurality of related information based on a previous recommendation list or a historical recommendation list operated by the user. It should be noted that the recommendation information already presented in the recommendation list is not included in the candidate recommendation information set.
In this embodiment, after the first vector and the second vector are determined, the click probability of each candidate recommendation information in the candidate recommendation information set may be determined according to the first vector and the second vector.
As a possible implementation manner, a pre-trained deep neural network model (DNN) may be adopted, the first vector and the second vector are input into the DNN model, and a vector corresponding to each candidate recommendation information in the candidate recommendation information set is input one by one, so as to obtain a click probability corresponding to each candidate recommendation information in the candidate recommendation information set.
And 104, selecting the (i + 1) th recommendation information from the candidate recommendation information set according to the click probability of each candidate recommendation information.
In this embodiment, after the click probability corresponding to each candidate recommendation information is determined, i +1 th recommendation information may be selected from the candidate recommendation information set according to the click probability corresponding to each candidate recommendation information, and the determined i +1 th recommendation information is displayed in the recommendation list. For example, the candidate recommendation information with the highest click probability is determined as the (i + 1) th recommendation information.
The recommendation information processing method of this embodiment is a repeatedly executed process, and after the (i + 1) th recommendation information is determined, the above process may be repeated, and the (i + 2) th recommendation information is continuously determined, and so on, until the display areas in the display list are all occupied or no candidate recommendation information exists in the candidate recommendation information set.
It should be noted that, when predicting the first recommendation information, the first recommendation information in the recommendation list may be determined in different manners.
As an example, when the recommendation list is empty (i.e., i ═ 0), two vectors may be randomly generated as the first vector and the second vector, respectively. Taking the example of determining the first vector by using the bag-of-words list and the one-hot encoding method, the first vector and the second vector may be randomly generated, wherein the dimensions of the first vector and the second vector are consistent with the length of the bag-of-words list. And then, according to the first vector and the second vector, determining the click probability of each candidate recommendation information in the candidate recommendation information set, and further determining the first recommendation information according to the click probability.
As an example, when the recommendation list is empty (i.e. i is 0), a conventional scoring model may be further used to score each candidate recommendation information in the candidate recommendation information set, and determine the candidate recommendation information with the highest score as the first recommendation information.
In the recommendation information processing method of this embodiment, a first vector corresponding to the ith recommendation information in the recommendation list is determined, a second vector corresponding to the first i recommendation information is determined according to the first vector, the click probability of each candidate recommendation information in the candidate recommendation information set is further determined according to the first vector and the second vector, and the (i + 1) th recommendation information is selected from the candidate recommendation information set according to the click probability of each candidate recommendation information. Therefore, the current recommendation information is determined according to the vector of the previous recommendation information and the vectors of all the recommendation information which are sequenced, the recommendation information is sequenced according to the relevance between the recommendation information, the click rate of the recommendation information in the recommendation list is improved, and the user experience is improved.
In practical applications, a user usually does not leave immediately after browsing currently recommended information, and in most cases, the user often interacts with a product (such as a browser) for many times to acquire more information when browsing information. All information browsing activities that occur during a period of time, such as half an hour, etc., from when a user enters a product to when the user leaves the product, may be referred to as a session. The information contained in the conversation can reflect the interest of the user in a short time, and is favorable for recommending the information to the user in a targeted manner. Based on the method, the application also provides another recommended information processing method. Fig. 3 is a flowchart illustrating another recommended information processing method according to an embodiment of the present application.
As shown in fig. 3, the recommended information processing method may include the steps of:
step 201, determining a first vector corresponding to the ith recommendation information in a recommendation list, wherein the recommendation list is the jth recommendation list corresponding to the recommendation object, and i and j are positive integers.
And the recommendation object is a user using a terminal corresponding to the recommendation list.
Step 202, according to the first vector, determining a second vector corresponding to the previous i pieces of recommendation information.
In this embodiment, for the description of step 201 to step 202, reference may be made to the description of step 101 to step 102 in the foregoing embodiment, and details are not described here again.
Step 203, determining a session vector corresponding to the recommended object currently.
The session refers to all information browsing behaviors of the recommended object in a certain time period, and can reflect the short-term interest of the recommended object. The conversation comprises information clicked and browsed by the recommendation object, vectors corresponding to the information clicked and browsed by the recommendation object can be determined, and then the conversation vector is determined according to the determined vectors.
As a possible implementation manner, when determining a session vector corresponding to a recommendation object currently, a pre-trained Recurrent Neural Networks (RNN) model may be adopted, click and browse information of a user on recommendation information in each recommendation list during a session is input into the RNN model to obtain a vector of each recommendation list, and a session vector corresponding to the recommendation object currently may be determined according to the vector of each recommendation list. The RNN model can adopt a one-way RNN model to describe the time sequence relation among all information in the conversation, so that the interest and the reading mode of the recommended object in a short term can be better described.
As a possible implementation manner, when determining the session vector corresponding to the current recommendation object, the click vector corresponding to each recommendation list in the previous j recommendation lists may be determined according to the click information corresponding to each recommendation information in the previous j recommendation lists, and then the current corresponding session vector may be determined according to the click vector corresponding to each recommendation list in the previous j recommendation lists and the recommendation order of the previous j recommendation lists. For example, for each recommendation list, the bag-of-word list and the one-hot coding method described in the foregoing embodiment may be adopted to determine a vector corresponding to each piece of recommendation information clicked by the user in the recommendation list, and then sum the obtained vectors to obtain a click vector corresponding to the recommendation list. And then, distributing a corresponding weight value for each recommendation list according to the recommendation sequence of the previous j recommendation lists, and determining the current corresponding session vector by adopting a weighted summation mode.
As a possible implementation manner, when determining the session vector corresponding to the current recommendation object, the vectors corresponding to all the clicked recommendation information of the users in the previous j recommendation lists may be determined, and then all the obtained vectors are summed to obtain the session vector corresponding to the current recommendation object.
And 204, determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the current corresponding session vector, the first vector and the second vector.
In this embodiment, after the session vector corresponding to the recommendation object is determined, the click probability of each candidate recommendation information in the candidate recommendation information set may be determined according to the session vector, the first vector, and the second vector corresponding to the recommendation object.
For example, the session vector, the first vector, and the second vector may be input into the DNN model, and a vector corresponding to each candidate recommendation information in the candidate recommendation information set is input one by one, so as to obtain a click probability corresponding to each candidate recommendation information in the candidate recommendation information set.
Step 205, according to the click probability of each candidate recommendation information, selecting the (i + 1) th recommendation information from the candidate recommendation information set.
For example, the candidate recommendation information with the highest click probability in the candidate recommendation information set may be determined as the (i + 1) th recommendation information.
According to the recommendation information processing method, the first vector corresponding to the ith recommendation information in the recommendation list, the second vector corresponding to the previous i recommendation information in the recommendation list and the conversation vector corresponding to the recommendation object are determined, the candidate recommendation information and the click probability of each candidate recommendation information are determined according to the conversation vector, the first vector and the second vector, and the (i + 1) th recommendation information is selected from the candidate recommendation information set according to the click probability, so that the relevance among the recommendation information and the interest of a user in a short time are considered, the recommendation information in the recommendation list can meet the current requirements of the user, and the user experience is further improved.
When the user is in different scenes, the browsed information contents are different. For example, in leisure time at night, the user browses more video resources, and under the condition of poor outdoor network conditions, the user browses more graphic and text information. Therefore, when information recommendation is carried out, the personalized requirements of the user in different scenes can be further met by combining the scene where the user is located. Based on this, the present application proposes another recommended information processing method, and fig. 4 is a flowchart illustrating another recommended information processing method provided in the embodiment of the present application.
As shown in fig. 4, the recommended information processing method may include a step of:
step 301, determining a first vector corresponding to the ith recommendation information in the recommendation list, where i is a positive integer.
Step 302, according to the first vector, determining a second vector corresponding to the first i pieces of recommendation information.
In this embodiment, for the description of steps 301 to 302, reference may be made to the description of steps 101 to 102 in the foregoing embodiment, and details are not described here again.
Step 303, determining a scene vector corresponding to the recommendation list.
The scene information includes, but is not limited to, time information, attribute information of the terminal device, network status, location information, refresh frequency, and the like.
When the recommended objects are in different scenes, the information browsed by the user is different. For example, when the network state of the terminal device used by the recommendation object is poor, the recommendation object tends to browse text information; when the network state of the terminal equipment is good and the recommended object is in a leisure state, the recommended object tends to browse the video information. Therefore, in this embodiment, recommendation information may be predicted in combination with current scene information, so that the recommendation information is matched with the current scene. Specifically, current scene information may be acquired, and then a scene vector corresponding to the recommendation list may be determined.
In a possible implementation manner of the embodiment of the application, when determining the scene vector corresponding to the recommendation list, the scene vector corresponding to the recommendation list may be determined according to at least one of time information, a refresh frequency, an attribute of the terminal device, a position of the terminal device, and a network state of the terminal device corresponding to the recommendation list.
As an example, the scene vector corresponding to the recommendation list may be determined according to a bag of words list used when determining the first vector and the second vector. When the scene information is multiple, a vector corresponding to each scene information may be determined, and then the multiple vectors are subjected to summation or averaging to obtain a scene vector.
As an example, possible scene information may be counted in advance, a corresponding vector may be set for each scene information, a mapping relationship between each scene information and the corresponding vector may be stored, and when a scene vector corresponding to the recommendation list is determined, the corresponding scene vector may be determined by querying the mapping relationship according to the scene information. When the scene information is multiple, a vector corresponding to each scene information may be determined, and then the multiple vectors are subjected to summation or averaging to obtain a scene vector.
And step 304, determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the scene vector, the first vector and the second vector.
In this embodiment, after the first vector, the second vector, and the scene vector are determined, the click probability of each candidate recommendation information in the candidate recommendation information set may be determined according to the first vector, the second vector, and the scene vector.
As a possible implementation manner, when determining the click probability of each candidate recommendation information in the candidate recommendation information set, the scene vector, the first vector, and the second vector may be input into a pre-trained DNN model, and the vector corresponding to each candidate recommendation information in the candidate recommendation information set is input one by one, so as to obtain the click probability corresponding to each candidate recommendation information in the candidate recommendation information set.
Further, in a possible implementation manner of the embodiment of the application, the pre-trained DNN model may be provided with two prediction layers, and when determining the click probability of each candidate recommendation information in the candidate recommendation information set, the primary description vector corresponding to each candidate recommendation information may be determined according to a third vector, a first vector, and a second vector corresponding to each recommendation information, where the third vector corresponding to each recommendation information may be obtained in the same manner as the first vector. And then, determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the primary description vector and the scene vector. Therefore, the current scene information is combined when the click probability of the candidate recommendation information is finally determined, so that the influence degree of the scene information on the prediction result is improved, and the recommendation information is more consistent with the current scene.
And 305, selecting the (i + 1) th recommendation information from the candidate recommendation information set according to the click probability of each candidate recommendation information.
In this embodiment, after the click probability of each candidate recommendation information is determined, the (i + 1) th recommendation information can be determined from the candidate recommendation information set according to the click probability of each candidate recommendation information.
According to the recommendation information processing method, the first vector corresponding to the ith recommendation information in the recommendation list, the second vector corresponding to the previous i recommendation information in the recommendation list and the scene vector corresponding to the recommendation list are determined, the candidate recommendation information and the click probability of each candidate recommendation information are determined according to the scene vector, the first vector and the second vector, and the (i + 1) th recommendation information is selected from the candidate recommendation information set according to the click probability.
Different users have different preferences for browsing information, for example, some users prefer to browse military information, and some users prefer to browse vehicle-related information. In order to realize targeted recommendation of information to users and meet personalized needs of different users, in a possible implementation manner of the embodiment of the application, information recommendation can be performed to the users in combination with related information of the users. Therefore, the present application provides another recommended information processing method, and fig. 5 is a flowchart illustrating another recommended information processing method provided in an embodiment of the present application.
As shown in fig. 5, the recommended information processing method may include the steps of:
step 401, determining a first vector corresponding to the ith recommendation information in the recommendation list, where i is a positive integer.
Step 402, according to the first vector, determining a second vector corresponding to the previous i pieces of recommendation information.
In this embodiment, for the description of steps 401 to 402, reference may be made to the description of steps 101 to 102 in the foregoing embodiment, and details are not described here again.
Step 403, determining the portrait vector of the recommendation object corresponding to the recommendation list.
Because the profession and the hobbies and interests of the recommended objects are different, the information concerned by different recommended objects is also different. In order to implement personalized information recommendation for different recommendation objects, in this embodiment, a recommendation list may also be determined according to information of the recommendation object. The information of the recommended object includes, but is not limited to, basic information of the recommended object such as age, gender, and occupation, and historical browsing information of the recommended object.
In this embodiment, information of the recommended object may be obtained first, and then the portrait vector of the recommended object corresponding to the recommendation list may be determined according to the information of the recommended object.
As an example, vectors corresponding to basic information of the recommendation object, such as gender, age, and interests, may be determined in advance, a mapping relationship between each piece of basic information and the corresponding vector may be established, and when the information of the recommendation object includes the basic information of the recommendation object, the vector corresponding to the basic information may be determined by querying the mapping relationship. When the information of the recommended object includes the historical browsing information of the recommended object, the vector corresponding to the historical browsing information may be determined in the same manner as the first vector is determined. Further, a vector corresponding to the basic information and a vector corresponding to the history browsing information are used together as an image vector to be recommended.
Step 404, determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the portrait vector, the first vector and the second vector of the recommendation object.
For example, when the click probability of each piece of candidate recommendation information in the candidate recommendation information set is determined, the image vector, the first vector and the second vector of the recommendation object may be input into a pre-trained DNN model, and the vector corresponding to each piece of candidate recommendation information in the candidate recommendation information set may be input one by one, so as to obtain the click probability corresponding to each piece of candidate recommendation information in the candidate recommendation information set.
Step 405, according to the click probability of each candidate recommendation information, selecting the (i + 1) th recommendation information from the candidate recommendation information set.
For example, the candidate recommendation information with the highest click probability in the candidate recommendation information set may be determined as the (i + 1) th recommendation information.
According to the recommendation information processing method, the first vector corresponding to the ith recommendation information and the second vector corresponding to the previous i recommendation information in the recommendation list are determined, the portrait vector of the recommendation object is determined, the candidate recommendation information and the click probability of each candidate recommendation information are determined according to the portrait vector, the first vector and the second vector, and the (i + 1) th recommendation information is selected from the candidate recommendation information set according to the click probability, so that the relevance among the recommendation information and the information of the user are considered, targeted information recommendation is realized, the individuation of information recommendation is improved, and the user experience is further improved.
It should be noted that the recommendation information processing method in the foregoing embodiment may be implemented alone or in combination, for example, information recommendation is performed in combination with user information and scenario information, or information recommendation is performed in combination with session information and scenario information. The foregoing embodiments are merely exemplary and are not to be construed as limiting the present application.
Fig. 6 is a diagram illustrating an architecture of a recommendation information ranking model for implementing the recommendation information processing method of the present application. As shown in fig. 6, the recommendation information ranking model includes two Fully connected layers and two Full Convolutional Network (FCN) layers, where the FCN layer is used to predict recommendation information and can receive an input of any size, so as to avoid the limitation of input size and improve flexibility. As can be seen from fig. 6, when predicting recommendation information, first, a user information vector, a candidate recommendation information vector, an ith recommendation information vector, an i-th recommendation information vector, and a session vector are input to a first full connection layer, and an output result of the first full connection layer is input to a first FCN layer, so as to determine initial recommendation information. The session vector can be determined by adopting a unidirectional RNN model according to the session information of the user in a short time. Through the first full-connection layer, information such as interest point distribution, resource types, presentation styles and the like contained in the session information can be learned, and personalized information of the user can be learned. Then, the determined initial recommendation information and the scene vector are input into a second full-connected layer, and the output result of the second full-connected layer is input into a second FCN layer, so that the click probability of the (i + 1) th recommendation information can be obtained. As shown in fig. 6, the recommendation information is predicted at the second full-link layer in combination with the scene information, so that the influence of the scene information on the prediction result is enhanced, and the personalized requirements of the user in different scenes are further reflected. When the recommendation information ranking model is used for ranking recommendation information, the relevance among the recommendation information is considered, the personalized requirements of users, the requirements under different scenes and the short-term interest are considered, the accuracy and click rate of recommendation information prediction are improved, and the user experience is improved.
In order to implement the above embodiments, the present application also provides a recommended information processing apparatus.
Fig. 7 is a schematic structural diagram of a recommended information processing apparatus according to an embodiment of the present application.
As shown in fig. 7, the recommended information processing apparatus 50 includes: a first determination module 510, a second determination module 520, a probability determination module 530, and a selection module 540. Wherein,
the first determining module 510 is configured to determine a first vector corresponding to the ith recommendation information in the recommendation list, where i is a positive integer.
And a second determining module 520, configured to determine, according to the first vector, a second vector corresponding to the first i pieces of recommendation information.
In a possible implementation manner of the embodiment of the present application, the second determining module 520 is specifically configured to determine a weight corresponding to each piece of recommendation information according to a position relationship between each piece of recommendation information and the (i + 1) th piece of recommendation information in a recommendation list; and determining a second vector corresponding to the first i pieces of recommendation information according to the weight and the vector corresponding to each piece of recommendation information.
A probability determining module 530, configured to determine, according to the first vector and the second vector, a click probability of each candidate piece of recommendation information in the candidate piece of recommendation information set.
The selecting module 540 is configured to select the (i + 1) th recommendation information from the candidate recommendation information set according to the click probability of each candidate recommendation information.
In a possible implementation manner of the embodiment of the application, the recommendation list is a jth recommendation list corresponding to the recommendation object, where j is a positive integer. As shown in fig. 8, on the basis of the embodiment shown in fig. 7, the recommendation information processing apparatus 50 may further include:
and a session vector determining module 550, configured to determine a session vector currently corresponding to the recommended object.
Specifically, the session vector determining module 550 is configured to determine, according to click information corresponding to each piece of recommendation information in the previous j recommendation lists, a click vector corresponding to each recommendation list in the previous j recommendation lists; and determining the current corresponding session vector according to the click vector corresponding to each recommendation list in the previous j recommendation lists and the recommendation sequence of the previous j recommendation lists.
Thus, in this embodiment, the probability determining module 530 is further configured to determine the click probability of each candidate recommendation information in the candidate recommendation information set according to the currently corresponding session vector, the first vector, and the second vector.
When the (i + 1) th recommendation information is determined, the click probability of the candidate recommendation information is determined by combining the session vector, and the (i + 1) th recommendation information is determined according to the click probability, so that not only is the relevance between the recommendation information considered, but also the interest of a user in a short time is considered, the recommendation information in the recommendation list can meet the current requirements of the user, and the user experience is further improved.
In a possible implementation manner of the embodiment of the present application, as shown in fig. 9, on the basis of the embodiment shown in fig. 7, the recommended information processing apparatus 50 may further include:
and a scene vector determining module 560, configured to determine a scene vector corresponding to the recommendation list.
Specifically, the scene vector determining module 560 is configured to determine the scene vector corresponding to the recommendation list according to at least one of time information, a refresh frequency, an attribute of the terminal device, a location of the terminal device, and a network state of the terminal device corresponding to the recommendation list.
Thus, in this embodiment, the probability determining module 530 is further configured to determine, according to the scene vector, the first vector, and the second vector, a click probability of each candidate recommendation information in the candidate recommendation information set.
Specifically, the probability determining module 530 is configured to determine a primary description vector corresponding to each candidate recommendation information according to the third vector, the first vector, and the second vector corresponding to each recommendation information; and determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the primary description vector and the scene vector.
The click probability of the candidate recommendation information is determined by combining the scene vectors corresponding to the recommendation list, and the (i + 1) th recommendation information is determined according to the click probability, so that when the recommendation information is predicted, not only the relevance among the recommendation information is considered, but also the current scene information is considered, the recommendation information in the recommendation list can meet the current requirements of the user, the recommendation list obtained by the user conforms to the current use scene, and the user experience is further improved.
In a possible implementation manner of the embodiment of the present application, as shown in fig. 10, on the basis of the embodiment shown in fig. 7, the recommended information processing apparatus 50 may further include:
and the portrait vector determining module 570 is used for determining the portrait vectors of the recommended objects corresponding to the recommendation list.
Therefore, in this embodiment, the probability determining module 530 is further configured to determine the click probability of each candidate recommendation information in the candidate recommendation information set according to the portrait vector, the first vector, and the second vector of the recommendation object.
The click probability of the candidate recommendation information is determined by combining the portrait vector of the recommendation object, and the (i + 1) th recommendation information is determined according to the click probability, so that when the recommendation information is predicted, not only the relevance among the recommendation information is considered, but also the information of the user is considered, targeted information recommendation is realized, the individuation of information recommendation is improved, and the user experience is further improved.
It should be noted that the foregoing explanation of the recommended information processing method embodiment is also applicable to the recommended information processing apparatus of this embodiment, and the implementation principle is similar, and is not described herein again.
According to the recommendation information processing device, the first vector corresponding to the ith recommendation information in the recommendation list is determined, the second vector corresponding to the previous i recommendation information is determined according to the first vector, the click probability of each candidate recommendation information in the candidate recommendation information set is further determined according to the first vector and the second vector, and the (i + 1) th recommendation information is selected from the candidate recommendation information set according to the click probability of each candidate recommendation information. Therefore, the current recommendation information is determined according to the vector of the previous recommendation information and the vectors of all the recommendation information which are sequenced, the recommendation information is sequenced according to the relevance between the recommendation information, the click rate of the recommendation information in the recommendation list is improved, and the user experience is improved.
In order to implement the foregoing embodiments, the present application also provides a computer device, including: a processor and a memory. Wherein the processor runs a program corresponding to the executable program code by reading the executable program code stored in the memory, for implementing the recommended information processing method as described in the foregoing embodiments.
FIG. 11 is a block diagram of a computer device provided in an embodiment of the present application, illustrating an exemplary computer device 90 suitable for use in implementing embodiments of the present application. The computer device 90 shown in fig. 11 is only an example, and should not bring any limitation to the function and the scope of use of the embodiments of the present application.
As shown in fig. 11, the computer device 90 is in the form of a general purpose computer device. The components of computer device 90 may include, but are not limited to: one or more processors or processing units 906, a system memory 910, and a bus 908 that couples the various system components (including the system memory 910 and the processing unit 906).
Bus 908 represents one or more of any of several types of bus structures, including a memory bus or memory controller, a peripheral bus, an accelerated graphics port, and a processor or local bus using any of a variety of bus architectures. These architectures include, but are not limited to, Industry Standard Architecture (ISA) bus, Micro Channel Architecture (MAC) bus, enhanced ISA bus, Video Electronics Standards Association (VESA) local bus, and Peripheral Component Interconnect (PCI) bus, to name a few.
Computer device 90 typically includes a variety of computer system readable media. Such media may be any available media that is accessible by computer device 90 and includes both volatile and nonvolatile media, removable and non-removable media.
The system Memory 910 may include computer system readable media in the form of volatile Memory, such as Random Access Memory (RAM) 911 and/or cache Memory 912. The computer device 90 may further include other removable/non-removable, volatile/nonvolatile computer system storage media. By way of example only, storage system 913 may be used to read from and write to non-removable, nonvolatile magnetic media (not shown in FIG. 11, and commonly referred to as a "hard disk drive"). Although not shown in FIG. 11, a disk drive for reading from and writing to a removable, nonvolatile magnetic disk (e.g., a "floppy disk") and an optical disk drive for reading from or writing to a removable, nonvolatile optical disk (e.g., a Compact disk Read Only Memory (CD-ROM), a Digital versatile disk Read Only Memory (DVD-ROM), or other optical media) may be provided. In these cases, each drive may be connected to bus 908 by one or more data media interfaces. System memory 910 may include at least one program product having a set (e.g., at least one) of program modules that are configured to carry out the functions of embodiments of the application.
A computer readable signal medium may include a propagated data signal with computer readable program code embodied therein, for example, in baseband or as part of a carrier wave. Such a propagated data signal may take any of a variety of forms, including, but not limited to, electro-magnetic, optical, or any suitable combination thereof. A computer readable signal medium may also be any computer readable medium that is not a computer readable storage medium and that can communicate, propagate, or transport a program for use by or in connection with an instruction execution system, apparatus, or device.
Program code embodied on a computer readable medium may be transmitted using any appropriate medium, including but not limited to wireless, wireline, optical fiber cable, RF, etc., or any suitable combination of the foregoing.
Computer program code for carrying out operations for aspects of the present application may be written in any combination of one or more programming languages, including an object oriented programming language such as Java, Smalltalk, C + +, and conventional procedural programming languages, such as the "C" programming language or similar programming languages. The program code may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer or entirely on the remote computer or server.
Program/utility 914 having a set (at least one) of program modules 9140 may be stored, for example, in system memory 910, such program modules 9140 including, but not limited to, an operating system, one or more application programs, other program modules, and program data, each of which or some combination of these examples may comprise an implementation of a network environment. Program modules 9140 generally perform the functions and/or methods of embodiments described herein.
The computer device 90 may also communicate with one or more external devices 10 (e.g., keyboard, pointing device, display 100, etc.), with one or more devices that enable a user to interact with the terminal device 90, and/or with any devices (e.g., network card, modem, etc.) that enable the computer device 90 to communicate with one or more other computing devices. Such communication may occur via input/output (I/O) interfaces 902. Moreover, computer device 90 may also communicate with one or more networks (e.g., a Local Area Network (LAN), a Wide Area Network (WAN), and/or a public Network such as the Internet) via Network adapter 900. As shown in FIG. 11, network adapter 900 communicates with the other modules of computer device 90 via bus 908. It should be appreciated that although not shown in FIG. 11, other hardware and/or software modules may be used in conjunction with computer device 90, including but not limited to: microcode, device drivers, redundant processing units, external disk drive arrays, RAID systems, tape drives, and data backup storage systems, among others.
The processing unit 906 executes various functional applications and data processing by executing programs stored in the system memory 910, for example, implementing the recommended information processing method mentioned in the foregoing embodiments.
In order to implement the above-mentioned embodiments, the present application also proposes a non-transitory computer-readable storage medium having stored thereon a computer program which, when executed by a processor, implements the recommended information processing method as described in the foregoing embodiments.
In order to implement the foregoing embodiments, the present application also proposes a computer program product, wherein when the instructions in the computer program product are executed by a processor, the recommended information processing method according to the foregoing embodiments is implemented.
In the description herein, reference to the description of the term "one embodiment," "some embodiments," "an example," "a specific example," or "some examples," etc., means that a particular feature, structure, material, or characteristic described in connection with the embodiment or example is included in at least one embodiment or example of the application. In this specification, the schematic representations of the terms used above are not necessarily intended to refer to the same embodiment or example. Furthermore, the particular features, structures, materials, or characteristics described may be combined in any suitable manner in any one or more embodiments or examples. Furthermore, various embodiments or examples and features of different embodiments or examples described in this specification can be combined and combined by one skilled in the art without contradiction.
Furthermore, the terms "first", "second" and "first" are used for descriptive purposes only and are not to be construed as indicating or implying relative importance or implicitly indicating the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one such feature. In the description of the present application, "plurality" means at least two, e.g., two, three, etc., unless specifically limited otherwise.
Any process or method descriptions in flow charts or otherwise described herein may be understood as representing modules, segments, or portions of code which include one or more executable instructions for implementing steps of a custom logic function or process, and alternate implementations are included within the scope of the preferred embodiment of the present application in which functions may be executed out of order from that shown or discussed, including substantially concurrently or in reverse order, depending on the functionality involved, as would be understood by those reasonably skilled in the art of the present application.
The logic and/or steps represented in the flowcharts or otherwise described herein, e.g., an ordered listing of executable instructions that can be considered to implement logical functions, can be embodied in any computer-readable medium for use by or in connection with an instruction execution system, apparatus, or device, such as a computer-based system, processor-containing system, or other system that can fetch the instructions from the instruction execution system, apparatus, or device and execute the instructions. For the purposes of this description, a "computer-readable medium" can be any means that can contain, store, communicate, propagate, or transport the program for use by or in connection with the instruction execution system, apparatus, or device. More specific examples (a non-exhaustive list) of the computer-readable medium would include the following: an electrical connection (electronic device) having one or more wires, a portable computer diskette (magnetic device), a Random Access Memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), an optical fiber device, and a portable compact disc read-only memory (CDROM). Additionally, the computer-readable medium could even be paper or another suitable medium upon which the program is printed, as the program can be electronically captured, via for instance optical scanning of the paper or other medium, then compiled, interpreted or otherwise processed in a suitable manner if necessary, and then stored in a computer memory.
It should be understood that portions of the present application may be implemented in hardware, software, firmware, or a combination thereof. In the above embodiments, the various steps or methods may be implemented in software or firmware stored in memory and executed by a suitable instruction execution system. If implemented in hardware, as in another embodiment, any one or combination of the following techniques, which are known in the art, may be used: a discrete logic circuit having a logic gate circuit for implementing a logic function on a data signal, an application specific integrated circuit having an appropriate combinational logic gate circuit, a Programmable Gate Array (PGA), a Field Programmable Gate Array (FPGA), or the like.
It will be understood by those skilled in the art that all or part of the steps carried by the method for implementing the above embodiments may be implemented by hardware related to instructions of a program, which may be stored in a computer readable storage medium, and when the program is executed, the program includes one or a combination of the steps of the method embodiments.
In addition, functional units in the embodiments of the present application may be integrated into one processing module, or each unit may exist alone physically, or two or more units are integrated into one module. The integrated module can be realized in a hardware mode, and can also be realized in a software functional module mode. The integrated module, if implemented in the form of a software functional module and sold or used as a stand-alone product, may also be stored in a computer readable storage medium.
The storage medium mentioned above may be a read-only memory, a magnetic or optical disk, etc. Although embodiments of the present application have been shown and described above, it is understood that the above embodiments are exemplary and should not be construed as limiting the present application, and that variations, modifications, substitutions and alterations may be made to the above embodiments by those of ordinary skill in the art within the scope of the present application.
Claims (11)
1. A recommended information processing method, characterized by comprising:
determining a first vector corresponding to ith recommendation information in a recommendation list, wherein i is a positive integer;
determining second vectors corresponding to the first i pieces of recommendation information according to the first vectors;
determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the first vector and the second vector;
and selecting the (i + 1) th recommendation information from the candidate recommendation information set according to the click probability of each candidate recommendation information.
2. The method of claim 1, wherein the determining the second vector corresponding to the first i pieces of recommendation information comprises:
determining a weight corresponding to each piece of recommendation information according to the position relation between each piece of recommendation information and the (i + 1) th piece of recommendation information in a recommendation list;
and determining a second vector corresponding to the first i pieces of recommendation information according to the weight and the vector corresponding to each piece of recommendation information.
3. The method of claim 1, wherein the recommendation list is a jth recommendation list corresponding to a recommended object, where j is a positive integer;
before determining the click probability of each candidate recommendation information in the candidate recommendation information set, the method further includes:
determining a session vector corresponding to the recommended object currently;
the determining the click probability of each candidate recommendation information in the candidate recommendation information set comprises:
and determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the current corresponding session vector, the first vector and the second vector.
4. The method of claim 3, wherein the determining the session vector to which the recommended object currently corresponds comprises:
determining a click vector corresponding to each recommendation list in the previous j recommendation lists according to the click information corresponding to each recommendation information in the previous j recommendation lists;
and determining a current corresponding session vector according to the click vector corresponding to each recommendation list in the previous j recommendation lists and the recommendation sequence of the previous j recommendation lists.
5. The method of claim 1, wherein prior to determining the probability of click for each candidate recommendation in the set of candidate recommendations, further comprising:
determining a scene vector corresponding to the recommendation list;
the determining the click probability of each candidate recommendation information in the candidate recommendation information set comprises:
and determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the scene vector, the first vector and the second vector.
6. The method of claim 5, wherein the determining the probability of click for each candidate recommendation in the set of candidate recommendations based on the scene vector, the first vector, and the first vector comprises:
determining a primary description vector corresponding to each candidate recommendation information according to the third vector, the first vector and the second vector corresponding to each recommendation information;
and determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the primary description vector and the scene vector.
7. The method of claim 5, wherein the determining the scene vector to which the recommendation list corresponds comprises:
and determining a scene vector corresponding to the recommendation list according to at least one of time information, refreshing frequency, attributes of the terminal equipment, the position of the terminal equipment and the network state of the terminal equipment corresponding to the recommendation list.
8. The method of any of claims 1-7, wherein prior to determining the probability of click for each candidate recommendation in the set of candidate recommendations, further comprising:
determining an portrait vector of a recommended object corresponding to the recommended list;
the determining the click probability of each candidate recommendation information in the candidate recommendation information set comprises:
and determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the portrait vector of the recommendation object, the first vector and the second vector.
9. A recommended information processing apparatus, characterized by comprising:
the device comprises a first determining module, a second determining module and a third determining module, wherein the first determining module is used for determining a first vector corresponding to the ith recommendation information in a recommendation list, and i is a positive integer;
the second determining module is used for determining second vectors corresponding to the first i pieces of recommendation information according to the first vectors;
the probability determining module is used for determining the click probability of each candidate recommendation information in the candidate recommendation information set according to the first vector and the second vector;
and the selection module is used for selecting the (i + 1) th recommendation information from the candidate recommendation information set according to the click probability of each candidate recommendation information.
10. A computer device comprising a processor and a memory;
wherein the processor executes a program corresponding to the executable program code by reading the executable program code stored in the memory, for implementing the recommended information processing method according to any one of claims 1 to 8.
11. A non-transitory computer-readable storage medium on which a computer program is stored, the program being characterized by implementing the recommended information processing method according to any one of claims 1 to 8 when executed by a processor.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810865257.1A CN109255070B (en) | 2018-08-01 | 2018-08-01 | Recommendation information processing method and device, computer equipment and storage medium |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810865257.1A CN109255070B (en) | 2018-08-01 | 2018-08-01 | Recommendation information processing method and device, computer equipment and storage medium |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109255070A true CN109255070A (en) | 2019-01-22 |
| CN109255070B CN109255070B (en) | 2022-04-12 |
Family
ID=65049034
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810865257.1A Active CN109255070B (en) | 2018-08-01 | 2018-08-01 | Recommendation information processing method and device, computer equipment and storage medium |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109255070B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112989276A (en) * | 2019-12-12 | 2021-06-18 | 北京沃东天骏信息技术有限公司 | Evaluation method and device of information push system |
| CN112989198A (en) * | 2021-03-30 | 2021-06-18 | 北京三快在线科技有限公司 | Push content determination method, device, equipment and computer-readable storage medium |
| CN113239175A (en) * | 2021-06-10 | 2021-08-10 | 中国平安人寿保险股份有限公司 | Method for displaying candidate sentence list and terminal equipment |
| CN114840762A (en) * | 2022-05-19 | 2022-08-02 | 马上消费金融股份有限公司 | Recommended content determining method and device and electronic equipment |
| WO2023001093A1 (en) * | 2021-07-22 | 2023-01-26 | 北京字节跳动网络技术有限公司 | Article recommendation method and apparatus, and device and storage medium |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103778260A (en) * | 2014-03-03 | 2014-05-07 | 哈尔滨工业大学 | Individualized microblog information recommending system and method |
| CN103885968A (en) * | 2012-12-20 | 2014-06-25 | 北京百度网讯科技有限公司 | Method and device for providing recommended information |
| CN104572734A (en) * | 2013-10-23 | 2015-04-29 | 腾讯科技（深圳）有限公司 | Question recommendation method, device and system |
| CN105678587A (en) * | 2016-01-12 | 2016-06-15 | 腾讯科技（深圳）有限公司 | Recommendation feature determination method and information recommendation method and device |
| CN106339510A (en) * | 2016-10-28 | 2017-01-18 | 北京百度网讯科技有限公司 | The click prediction method and device based on artificial intelligence |
| CN107424043A (en) * | 2017-06-15 | 2017-12-01 | 北京三快在线科技有限公司 | A kind of Products Show method and device, electronic equipment |
| CN107463701A (en) * | 2017-08-15 | 2017-12-12 | 北京百度网讯科技有限公司 | Method and apparatus based on artificial intelligence pushed information stream |
-
2018
- 2018-08-01 CN CN201810865257.1A patent/CN109255070B/en active Active
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103885968A (en) * | 2012-12-20 | 2014-06-25 | 北京百度网讯科技有限公司 | Method and device for providing recommended information |
| CN104572734A (en) * | 2013-10-23 | 2015-04-29 | 腾讯科技（深圳）有限公司 | Question recommendation method, device and system |
| CN103778260A (en) * | 2014-03-03 | 2014-05-07 | 哈尔滨工业大学 | Individualized microblog information recommending system and method |
| CN105678587A (en) * | 2016-01-12 | 2016-06-15 | 腾讯科技（深圳）有限公司 | Recommendation feature determination method and information recommendation method and device |
| CN106339510A (en) * | 2016-10-28 | 2017-01-18 | 北京百度网讯科技有限公司 | The click prediction method and device based on artificial intelligence |
| CN107424043A (en) * | 2017-06-15 | 2017-12-01 | 北京三快在线科技有限公司 | A kind of Products Show method and device, electronic equipment |
| CN107463701A (en) * | 2017-08-15 | 2017-12-12 | 北京百度网讯科技有限公司 | Method and apparatus based on artificial intelligence pushed information stream |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112989276A (en) * | 2019-12-12 | 2021-06-18 | 北京沃东天骏信息技术有限公司 | Evaluation method and device of information push system |
| CN112989198A (en) * | 2021-03-30 | 2021-06-18 | 北京三快在线科技有限公司 | Push content determination method, device, equipment and computer-readable storage medium |
| CN113239175A (en) * | 2021-06-10 | 2021-08-10 | 中国平安人寿保险股份有限公司 | Method for displaying candidate sentence list and terminal equipment |
| WO2023001093A1 (en) * | 2021-07-22 | 2023-01-26 | 北京字节跳动网络技术有限公司 | Article recommendation method and apparatus, and device and storage medium |
| CN114840762A (en) * | 2022-05-19 | 2022-08-02 | 马上消费金融股份有限公司 | Recommended content determining method and device and electronic equipment |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109255070B (en) | 2022-04-12 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109255070B (en) | Recommendation information processing method and device, computer equipment and storage medium | |
| CN110413877A (en) | A kind of resource recommendation method, device and electronic equipment | |
| US8694890B2 (en) | Use of color in a site analysis report | |
| CN112733042B (en) | Recommendation information generation method, related device and computer program product | |
| CN108491540B (en) | Text information pushing method and device and intelligent terminal | |
| US8250454B2 (en) | Client-side composing/weighting of ads | |
| CN104782138A (en) | Identifying a thumbnail image to represent a video | |
| JP5384452B2 (en) | Distribution server and recommendation method | |
| US20090319940A1 (en) | Network of trust as married to multi-scale | |
| CN112269917B (en) | Media resource display method and device, equipment, system and storage medium | |
| US8812957B2 (en) | Relevance slider in a site analysis report | |
| CN109074359B (en) | Optimizing content distribution using models | |
| US20200356781A1 (en) | Method and apparatus for performing categorised matching of videos, and selection engine | |
| US20080183860A1 (en) | Intelligent node positioning in a site analysis report | |
| CN108566615A (en) | Information-pushing method, device and client | |
| CN112241327A (en) | Shared information processing method and device, storage medium and electronic equipment | |
| US20200034893A1 (en) | Social infusion of media content | |
| CN111752426A (en) | Chat thread display method, recording medium, and computer device | |
| CN113159870B (en) | Push information display method and device and computer equipment | |
| CN108182240A (en) | The new gaining rate prediction model training of point of interest and Forecasting Methodology, device and storage medium | |
| CN114169418B (en) | Label recommendation model training method and device and label acquisition method and device | |
| CN115048577A (en) | Model training method, device, equipment and storage medium | |
| CN112948602B (en) | Content display method, device, system, equipment and storage medium | |
| CN113806631A (en) | Recommendation method, training method, device, equipment and news recommendation system | |
| US10841663B2 (en) | Methods and systems for managing content on multiple devices |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |