CN110245334A - Method and apparatus for output information - Google Patents
Method and apparatus for output information Download PDFInfo
- Publication number
- CN110245334A CN110245334A CN201910552619.6A CN201910552619A CN110245334A CN 110245334 A CN110245334 A CN 110245334A CN 201910552619 A CN201910552619 A CN 201910552619A CN 110245334 A CN110245334 A CN 110245334A
- Authority
- CN
- China
- Prior art keywords
- word
- probability
- article
- sentence
- sample
- 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
- 238000000034 method Methods 0.000 title claims abstract description 66
- 238000005520 cutting process Methods 0.000 claims abstract description 32
- 238000012549 training Methods 0.000 claims description 51
- 238000005457 optimization Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 6
- 238000012545 processing Methods 0.000 description 10
- 238000010586 diagram Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 8
- 230000015654 memory Effects 0.000 description 6
- 238000006243 chemical reaction Methods 0.000 description 5
- 238000004891 communication Methods 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 5
- 230000007774 longterm Effects 0.000 description 4
- 230000003287 optical effect Effects 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000006835 compression Effects 0.000 description 2
- 238000007906 compression Methods 0.000 description 2
- 230000005291 magnetic effect Effects 0.000 description 2
- 206010054949 Metaplasia Diseases 0.000 description 1
- 238000013528 artificial neural network Methods 0.000 description 1
- 230000006399 behavior Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000005611 electricity Effects 0.000 description 1
- 239000000835 fiber Substances 0.000 description 1
- 238000003780 insertion Methods 0.000 description 1
- 230000037431 insertion Effects 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 210000003127 knee Anatomy 0.000 description 1
- 239000004973 liquid crystal related substance Substances 0.000 description 1
- 230000007787 long-term memory Effects 0.000 description 1
- 230000015689 metaplastic ossification Effects 0.000 description 1
- 210000005036 nerve Anatomy 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000005498 polishing Methods 0.000 description 1
- 230000000644 propagated effect Effects 0.000 description 1
- 230000000306 recurrent effect Effects 0.000 description 1
- 238000009877 rendering Methods 0.000 description 1
- 230000011218 segmentation Effects 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
- 230000006403 short-term memory Effects 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000013518 transcription Methods 0.000 description 1
- 230000035897 transcription Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/50—Information retrieval; Database structures therefor; File system structures therefor of still image data
- G06F16/58—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
- G06F16/5866—Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually using information manually generated, e.g. tags, keywords, comments, manually generated location and time information
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/12—Use of codes for handling textual entities
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/12—Use of codes for handling textual entities
- G06F40/137—Hierarchical processing, e.g. outlines
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/10—Text processing
- G06F40/189—Automatic justification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/205—Parsing
- G06F40/216—Parsing using statistical methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F40/00—Handling natural language data
- G06F40/20—Natural language analysis
- G06F40/279—Recognition of textual entities
- G06F40/289—Phrasal analysis, e.g. finite state techniques or chunking
-
- G—PHYSICS
- G10—MUSICAL INSTRUMENTS; ACOUSTICS
- G10L—SPEECH ANALYSIS TECHNIQUES OR SPEECH SYNTHESIS; SPEECH RECOGNITION; SPEECH OR VOICE PROCESSING TECHNIQUES; SPEECH OR AUDIO CODING OR DECODING
- G10L15/00—Speech recognition
- G10L15/26—Speech to text systems
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Audiology, Speech & Language Pathology (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Library & Information Science (AREA)
- Databases & Information Systems (AREA)
- Probability & Statistics with Applications (AREA)
- Human Computer Interaction (AREA)
- Acoustics & Sound (AREA)
- Multimedia (AREA)
- Machine Translation (AREA)
Abstract
Embodiment of the disclosure discloses the method and apparatus for output information.One specific embodiment of this method includes: to obtain audio-frequency information to be converted；Audio-frequency information is converted into text information；Text information is subjected to word cutting, obtains word sequence；For the word in word sequence, the word that probabilistic model obtains is connected by word trained in advance and connects the connection probability that probability tables inquire the word and connect probability and the word and all kinds of punctuates between next word of the word, and the linking objective based on the connection determine the probability word inquired；Word each in word sequence is connected article of the generation with punctuate with corresponding linking objective to export.Audio can be changed into automatically the article with punctuate by the embodiment.
Description
Technical field
Embodiment of the disclosure is related to field of computer technology, and in particular to the method and apparatus for output information.
Background technique
Field is automatically generated in article, the article that multimedia transcription automatically generates is also fewer, is according to structuring mostly
Text data generate article, this makes data source single, and the article of generation is not abundant enough, extensively；And human-edited
Multimedia article is again very time-consuming and cumbersome, causes the expense of unnecessary manpower, financial resources.Conventional method is mainly artificial
Editor, by manually converting text for related audio, then searches picture concerned according to audio theme etc., finally on network
Artificial text and picture by after conversion renders.
It is based on artificial method main problem: (1) for the conversion of audio: it is time-consuming and laborious by the way of artificial,
Accuracy rate is also not necessarily high；(2) selection of figure: relevant picture is selected according to theme, the mode of manual search will expend big
Measure manpower；(3) the tissue rendering of article, ultimately generates a readability strong article for related text and picture tissue.
Summary of the invention
Embodiment of the disclosure proposes the method and apparatus for output information.
In a first aspect, embodiment of the disclosure provides a kind of method for output information, comprising: obtain to be converted
Audio-frequency information；Audio-frequency information is converted into text information；Text information is subjected to word cutting, obtains word sequence；For in word sequence
Word, by word trained in advance connect the word that probabilistic model obtains connect probability tables inquire the word and with next word of the word
Between connect the connection probability of probability and the word and all kinds of punctuates, and the connection based on the connection determine the probability word inquired
Target；Word each in word sequence is connected article of the generation with punctuate with corresponding linking objective to export.
In some embodiments, word connection probability tables are obtained through the following steps: obtaining training sample set, training
Sample includes the sentence containing punctuate；Using the sentence of the training sample in training sample set as the input of LSTM model, instruction
Get word connection probabilistic model；It is general between each word and word according to being obtained in the pilot process of word connection probabilistic model training
Probability between rate and each word and each punctuate generates word and connects probability tables.
In some embodiments, training sample set is obtained, comprising: sample article is obtained, sample article is big by one
The granularity of sentence carries out cutting and obtains sample sentence set, wherein big sentence refers to the sentence to end up with fullstop, question mark or exclamation mark；It is right
Sample sentence in sample sentence set generates term vector as training sample after the sentence is carried out word cutting.
In some embodiments, this method further include: article is divided at least one paragraph.
In some embodiments, this method further include: determine the theme and entity of article；Obtain the theme and reality with article
The matched image of body；Graph text information is generated according to image and article.
In some embodiments, this method further include: graph text information is subjected to typesetting optimization.
Second aspect, embodiment of the disclosure provide a kind of device for output information, comprising: acquiring unit, quilt
It is configured to obtain audio-frequency information to be converted；Converting unit is configured to audio-frequency information being converted into text information；Word cutting list
Member is configured to text information carrying out word cutting, obtains word sequence；Judging unit is configured to lead to the word in word sequence
The word is inquired after the word connection probability tables that word connection probabilistic model trained in advance obtains and is connected between next word of the word
Connect the connection probability of probability and the word and all kinds of punctuates, and the linking objective based on the connection determine the probability word inquired；
Connection unit is configured to word each in word sequence connecting article progress of the generation with punctuate with corresponding linking objective defeated
Out.
In some embodiments, which further includes training unit, is configured to: obtaining training sample set, training sample
It originally include the sentence containing punctuate；Using the sentence of the training sample in training sample set as the input of LSTM model, training
Obtain word connection probabilistic model；It is connected according to word in the pilot process of probabilistic model training and obtains the probability between each word and word
Probability between each word and each punctuate generates word and connects probability tables.
In some embodiments, training unit is further configured to: sample article is obtained, sample article is big by one
The granularity of sentence carries out cutting and obtains sample sentence set, wherein big sentence refers to the sentence to end up with fullstop, question mark or exclamation mark；It is right
Sample sentence in sample sentence set generates term vector as training sample after the sentence is carried out word cutting.
In some embodiments, which further includes segmenting unit, is configured to: article is divided at least one paragraph.
In some embodiments, which further includes figure unit, is configured to: determining the theme and entity of article；It obtains
Take the image with the theme of article and Entities Matching；Graph text information is generated according to image and article.
In some embodiments, which further includes typesetting unit, is configured to: graph text information is carried out typesetting optimization.
The third aspect, embodiment of the disclosure provide a kind of electronic equipment, comprising: one or more processors；Storage
Device is stored thereon with one or more programs, when one or more programs are executed by one or more processors, so that one
Or multiple processors are realized such as method any in first aspect.
Fourth aspect, embodiment of the disclosure provide a kind of computer-readable medium, are stored thereon with computer program,
Wherein, it realizes when program is executed by processor such as method any in first aspect.
The method and apparatus for output information that embodiment of the disclosure provides, the text that can be parsed according to audio
Content carries out sentence link and is segmented, and then carries out figure according to this paper subject content, finally excellent to text, picture progress typesetting
Metaplasia is at article.System is generated compared to traditional article, the data of the system are richer, various, and source is also more extensive.Compared to biography
The hand-written article of small volume of system, has higher timeliness and coverage, while also saving human cost and time cost.
Detailed description of the invention
By reading a detailed description of non-restrictive embodiments in the light of the attached drawings below, the disclosure is other
Feature, objects and advantages will become more apparent upon:
Fig. 1 is that one embodiment of the disclosure can be applied to exemplary system architecture figure therein；
Fig. 2 is the flow chart according to one embodiment of the method for output information of the disclosure；
Fig. 3 is the schematic diagram according to an application scenarios of the method for output information of the disclosure；
Fig. 4 is the flow chart according to another embodiment of the method for output information of the disclosure；
Fig. 5 a, 5b are the schematic network structures according to the LSTM model of the method for output information of the disclosure.
Fig. 6 is the structural schematic diagram according to one embodiment of the device for output information of the disclosure；
Fig. 7 is adapted for the structural schematic diagram for the computer system for realizing the electronic equipment of embodiment of the disclosure.
Specific embodiment
The disclosure is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining related invention, rather than the restriction to the invention.It also should be noted that in order to
Convenient for description, part relevant to related invention is illustrated only in attached drawing.
It should be noted that in the absence of conflict, the feature in embodiment and embodiment in the disclosure can phase
Mutually combination.The disclosure is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
Fig. 1 is shown can be using the method for output information of the disclosure or the implementation of the device for output information
The exemplary system architecture 100 of example.
As shown in Figure 1, system architecture 100 may include terminal device 101,102,103, network 104 and server 105.
Network 104 between terminal device 101,102,103 and server 105 to provide the medium of communication link.Network 104 can be with
Including various connection types, such as wired, wireless communication link or fiber optic cables etc..
User can be used terminal device 101,102,103 and be interacted by network 104 with server 105, to receive or send out
Send message etc..Various telecommunication customer end applications can be installed, such as audio conversion text is answered on terminal device 101,102,103
With, web browser applications, shopping class application, searching class application, instant messaging tools, mailbox client, social platform software
Deng.
Terminal device 101,102,103 can be hardware, be also possible to software.When terminal device 101,102,103 is hard
When part, it can be with microphone, display screen and the various electronic equipments for supporting audio conversion text, including but not limited to intelligently
Mobile phone, tablet computer, E-book reader, MP3 player (Moving Picture Experts Group Audio Layer
III, dynamic image expert's compression standard audio level 3), MP4 (Moving Picture Experts Group Audio
Layer IV, dynamic image expert's compression standard audio level 4) player, pocket computer on knee and desktop computer etc.
Deng.When terminal device 101,102,103 is software, may be mounted in above-mentioned cited electronic equipment.It may be implemented
At multiple softwares or software module (such as providing Distributed Services), single software or software module also may be implemented into.
It is not specifically limited herein.
Server 105 can be to provide the server of various services, such as to showing on terminal device 101,102,103
Text provides the backstage editing server supported.Backstage editing server the data such as the audio received such as can analyze
Processing, and processing result (such as the article generated according to audio) is fed back into terminal device.
It should be noted that server can be hardware, it is also possible to software.When server is hardware, may be implemented
At the distributed server cluster that multiple servers form, individual server also may be implemented into.It, can when server is software
It, can also be with to be implemented as multiple softwares or software module (such as providing multiple softwares of Distributed Services or software module)
It is implemented as single software or software module.It is not specifically limited herein.
It should be noted that the method provided by embodiment of the disclosure for output information can be by terminal device
101, it 102,103 executes, can also be executed by server 105.Correspondingly, it can be set for the device of output information in terminal
In equipment 101,102,103, also it can be set in server 105.It is not specifically limited herein.
It should be understood that the number of terminal device, network and server in Fig. 1 is only schematical.According to realization need
It wants, can have any number of terminal device, network and server.
With continued reference to Fig. 2, the process of one embodiment of the method for output information according to the disclosure is shown
200.This is used for the method for output information, comprising the following steps:
Step 201, audio-frequency information to be converted is obtained.
In the present embodiment, can lead to for the executing subject of the method for output information (such as server shown in FIG. 1)
It crosses wired connection mode or radio connection and carries out the terminal reception audio-frequency information that voice is write using it from user.Audio
Information can be the audio file of various formats.It includes a large amount of sentences.It may include this section audio in the title of audio file
Title.
Step 202, audio-frequency information is converted into text information.
In the present embodiment, existing automatic speech recognition (ASR, Automatic Speech can be passed through
Recognition) audio-frequency information is converted into whole section of text by technology.It is one section based on the text that ASR is parsed not make pauses in reading unpunctuated ancient writings
Text stamp punctuate so also needing that it is cut and is linked according to the meaning of one's words.
Step 203, text information is subjected to word cutting, obtains word sequence.
In the present embodiment, whole section of text is subjected to word cutting operation based on the lexical structure of Chinese or English, obtained whole
The word sequence of section audio.Word cutting method may include the common word cutting modes such as maximum reverse matching method.It can first identify the language of audio
Kind, for example, Chinese, English or other languages.Then word cutting operation is carried out according to the lexical structure of the languages.
Step 204, for the word in word sequence, the word that probabilistic model obtains is connected by word trained in advance and connects probability
Table inquires the word and connects the connection probability of probability and the word and all kinds of punctuates between next word of the word, and based on looking into
The linking objective for the connection determine the probability word ask.
In the present embodiment, according to word connect probabilistic model generate word connect probability tables, for each word calculate itself and
The probability of next word and all kinds of punctuates takes the maximum word of probability value or punctuate to be linked.Word connects probability tables and is used for table
Levy the probability of word and word or all kinds of punctuates.We regard each word as independent, that is to say, that each word back is likely to
In addition punctuate (,.?！；Deng).For current word, the probability of the word and various punctuates and next word is calculated separately, is finally taken
The highest word of probability is linked.If probability is up to next word, illustrate here it is not necessary to which punctuate connection, directly carries out
Word addition is all right.If highest probability is punctuate, in the additional punctuation mark in the word back.All words are as above walked
Suddenly, the sentence connected with punctuate is finally obtained.For example, word sequence " I " " love " " China " " because ", successively inquire " I " with
Connection probability between " love ", and the connection probability of the punctuates such as " I " and fullstop, comma.Between available " me " and " love "
Maximum probability is connected, punctuate is not therefore used between " I " " love ".And the connection probability of " China " and fullstop is much larger than " China "
With " because ", fullstop is added below also greater than the connection probability of " China " and other punctuates, therefore in " China ".Word connects probability
Model is an emphasis of this subsystem, needs to train correlation model, obtains the probability occurred between word and word to generate word
Probability tables are connected, linking objective of the highest word of probability as the word is then taken.The generating process that word connects probability tables will be in step
It is introduced in rapid 401-403.
Step 205, word each in word sequence and corresponding linking objective are connected generate the article with punctuate carry out it is defeated
Out.
In the present embodiment, according to step 204 as a result, word each in word sequence is connected with corresponding linking objective
The article with punctuate is generated to be exported.
In some optional implementations of the present embodiment, this method further include: article is divided at least one paragraph.
Semantic analysis can be carried out to article, be then segmented article according to semanteme.The word content of identical semanteme is classified as one section.
In some optional implementations of the present embodiment, this method further include: determine the theme and entity of article；It obtains
Take the image with the theme of article and Entities Matching；Graph text information is generated according to image and article.It is obtained according to audio conversion module
The text data arrived, (entity is more fine-grained to the entity for excavating in text here, including personage such as star, things
Such as bank) and theme (classifications such as finance and economics, amusement, sport), then according to entity go sterogram library searching related entities figure,
Thematic map library searching related subject figure is removed according to theme.These pictures are text picture concerned, can be directly with making an issue of
Figure.
In some optional implementations of the present embodiment, this method further include: graph text information is subjected to typesetting optimization.
Automatically picture is inserted into the more reasonable position of article, and adjusts dimension of picture, so that the area ratio of word content and picture
Example reaches predetermined value.
With continued reference to the signal that Fig. 3, Fig. 3 are according to the application scenarios of the method for output information of the present embodiment
Figure.In the application scenarios of Fig. 3, server receives the audio file " Dali scene " of terminal transmission.User is in audio text
The local conditions and customs of Dali is described with voice in part.By ASR technology, audio file is parsed into whole section of text.So
Afterwards by after whole section of text word cutting, connection probability between query word and word, between word and each punctuate, by the word of maximum probability or
Punctuate is as linking objective.Each word generates the article with punctuate after being attached.Also picture inspection can be carried out according to content of text
Rope finds suitable figure.Then article is segmented according to text semantic.Finally the picture searched is inserted into article again
Carry out polishing processing.
The method provided by the above embodiment of the disclosure can carry out sentence chain according to the content of text that audio parses
& segmentation is connect, figure is then carried out according to this paper subject content, typesetting optimization finally is carried out to text, picture and generates article.It compares
System is generated in traditional article, the data of the system are richer, various, and source is also more extensive；Compared to traditional hand-written text of small volume
Chapter has higher timeliness and coverage, while also saving human cost and time cost.
With further reference to Fig. 4, it illustrates the processes 400 of another embodiment of the method for output information.The use
In the process 400 of the method for output information, comprising the following steps:
Step 401, training sample set is obtained.
In the present embodiment, can lead to for the executing subject of the method for output information (such as server shown in FIG. 1)
It crosses wired connection mode or radio connection obtains training sample set, wherein training sample includes the sentence containing punctuate
Son.The executing subject of process 400 can be identical as the executing subject of process 200, can also be different executing subjects.It can be taken by third party
It is engaged in generating word connection probability tables after device executes process 400, then issues the executing subject use of process 200.
Use normal newsletter archive or article as training data.
Firstly, carrying out article cuts sentence, article is subjected to cutting according to the granularity of one big sentence, big sentence refers to fullstop, asks
Number, exclamation mark ending sentence.Each big sentence is as a data；
Then, sentence word cutting is carried out, word cutting is carried out to sentence according to the lexical structure of English or Chinese；
Finally, carrying out word encode (coding), say that each word is embedding (insertion) and obtains each sentence
Embedding indicates to get training sample has been arrived.Here word includes punctuation mark.
Step 402, using the sentence of the training sample in training sample set as the input of LSTM model, training obtains word
Connect probabilistic model.
In the present embodiment, LSTM (Long Short-Term Memory) is shot and long term memory network, is a kind of time
Recognition with Recurrent Neural Network is suitable for being spaced and postpone relatively long critical event in processing and predicted time sequence.Original RNN's
Only one state (Fig. 5 a) of hidden layer, it is very sensitive for short-term input.So, if we are further added by a state
(Fig. 5 b) allows it to save long-term state.
LSTM is equally such structure, but duplicate module possesses a different structure.Different from single nerve
Network layer, be here there are four, interacted in a kind of very special mode.
In t moment, there are three the inputs of LSTM: the input value of current time network, the output valve of last moment LSTM,
And the location mode of last moment；There are two the outputs of LSTM: the cell-like of current time LSTM output valve and current time
State.
The key of LSTM is exactly how to control long term state.Herein, the thinking of LSTM is opened using three controls
It closes.First switch, responsible control continue to save long term state；Second switch is responsible for control immediate status and is input to length
Phase state；Third switch, is responsible for controlling whether using long term state as the output of current LSTM.
Step 403, connected according to word obtained in the pilot process of probabilistic model training probability between each word and word and
Probability between each word and each punctuate generates word and connects probability tables.
In the present embodiment, using the sentence after embedding as the input of LSTM model, training pattern.Pull model
Pilot process, obtain the connection probability between each word and word.It obtains the connection probability between each word is for statistical analysis
Word connects probability tables.Probability tables are connected by looking into word, the connection probability before word and word can be obtained.
With further reference to Fig. 6, as the realization to method shown in above-mentioned each figure, present disclose provides one kind for exporting letter
One embodiment of the device of breath, the Installation practice is corresponding with embodiment of the method shown in Fig. 2, which can specifically answer
For in various electronic equipments.
As shown in fig. 6, the device 600 for output information of the present embodiment includes: acquiring unit 601, converting unit
602, word cutting unit 603, judging unit 604 and connection unit 605.Wherein, acquiring unit 601 are configured to obtain to be converted
Audio-frequency information；Converting unit 602 is configured to audio-frequency information being converted into text information；Word cutting unit 603, is configured to
Text information is subjected to word cutting, obtains word sequence；Judging unit 604 is configured to for the word in word sequence, by instructing in advance
The word connection probability tables that experienced word connection probabilistic model obtains inquire the word and connected between next word of the word probability and
The connection probability of the word and all kinds of punctuates, and the linking objective based on the connection determine the probability word inquired；Connection unit
605, it is configured to connect word each in word sequence and corresponding linking objective and generates the article with punctuate and export.
In the present embodiment, for the acquiring unit 601 of the device of output information 600, converting unit 602, word cutting unit
603, the specific processing of judging unit 604 and connection unit 605 can be with reference to step 201, the step in Fig. 2 corresponding embodiment
202, step 203, step 204 and step 205.
In some optional implementations of the present embodiment, device 600 further includes training unit (attached to be not shown in the figure),
It is configured to: obtaining training sample set, wherein training sample includes the sentence containing punctuate；It will be in training sample set
Input of the sentence of training sample as LSTM model, training obtain word connection probabilistic model；Probabilistic model instruction is connected according to word
The probability between the probability and each word and each punctuate between each word and word is obtained in experienced pilot process generates word connection generally
Rate table.
In some optional implementations of the present embodiment, training unit is further configured to: sample article is obtained,
Sample article is subjected to cutting by the granularity of one big sentence and obtains sample sentence set, wherein big sentence refers to fullstop, question mark or sense
The sentence of exclamation ending；For the sample sentence in sample sentence set, term vector is generated after which is carried out word cutting as training sample
This.
In some optional implementations of the present embodiment, device 600 further includes segmenting unit (attached to be not shown in the figure),
It is configured to: article is divided at least one paragraph.
In some optional implementations of the present embodiment, device 600 further includes figure unit (attached to be not shown in the figure),
It is configured to: determining the theme and entity of article；It obtains and the theme of article and the image of Entities Matching；According to image and article
Generate graph text information.
In some optional implementations of the present embodiment, device 600 further includes typesetting unit (attached to be not shown in the figure),
It is configured to: graph text information is subjected to typesetting optimization.
Below with reference to Fig. 7, it illustrates the electronic equipment that is suitable for being used to realize embodiment of the disclosure, (example is as shown in figure 1
Server) 700 structural schematic diagram.Server shown in Fig. 7 is only an example, should not be to the function of embodiment of the disclosure
Any restrictions can be brought with use scope.
As shown in fig. 7, electronic equipment 700 may include processing unit (such as central processing unit, graphics processor etc.)
701, random access can be loaded into according to the program being stored in read-only memory (ROM) 702 or from storage device 708
Program in memory (RAM) 703 and execute various movements appropriate and processing.In RAM 703, it is also stored with electronic equipment
Various programs and data needed for 700 operations.Processing unit 701, ROM 702 and RAM 703 pass through the phase each other of bus 704
Even.Input/output (I/O) interface 705 is also connected to bus 704.
In general, following device can connect to I/O interface 705: including such as touch screen, touch tablet, keyboard, mouse, taking the photograph
As the input unit 706 of head, microphone, accelerometer, gyroscope etc.；Including such as liquid crystal display (LCD), loudspeaker, vibration
The output device 707 of dynamic device etc.；Storage device 708 including such as tape, hard disk etc.；And communication device 709.Communication device
709, which can permit electronic equipment 700, is wirelessly or non-wirelessly communicated with other equipment to exchange data.Although Fig. 7 shows tool
There is the electronic equipment 700 of various devices, it should be understood that being not required for implementing or having all devices shown.It can be with
Alternatively implement or have more or fewer devices.Each box shown in Fig. 7 can represent a device, can also root
According to needing to represent multiple devices.
Particularly, in accordance with an embodiment of the present disclosure, it may be implemented as computer above with reference to the process of flow chart description
Software program.For example, embodiment of the disclosure includes a kind of computer program product comprising be carried on computer-readable medium
On computer program, which includes the program code for method shown in execution flow chart.In such reality
It applies in example, which can be downloaded and installed from network by communication device 709, or from storage device 708
It is mounted, or is mounted from ROM 702.When the computer program is executed by processing unit 701, the implementation of the disclosure is executed
The above-mentioned function of being limited in the method for example.It should be noted that computer-readable medium described in embodiment of the disclosure can be with
It is computer-readable signal media or computer readable storage medium either the two any combination.It is computer-readable
Storage medium for example may be-but not limited to-the system of electricity, magnetic, optical, electromagnetic, infrared ray or semiconductor, device or
Device, or any above combination.The more specific example of computer readable storage medium can include but is not limited to: have
The electrical connection of one or more conducting wires, portable computer diskette, hard disk, random access storage device (RAM), read-only memory
(ROM), erasable programmable read only memory (EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-
ROM), light storage device, magnetic memory device or above-mentioned any appropriate combination.In embodiment of the disclosure, computer
Readable storage medium storing program for executing can be any tangible medium for including or store program, which can be commanded execution system, device
Either device use or in connection.And in embodiment of the disclosure, computer-readable signal media may include
In a base band or as the data-signal that carrier wave a part is propagated, wherein carrying computer-readable program code.It is this
The data-signal of propagation can take various forms, including but not limited to electromagnetic signal, optical signal or above-mentioned any appropriate
Combination.Computer-readable signal media can also be any computer-readable medium other than computer readable storage medium, should
Computer-readable signal media can send, propagate or transmit for by instruction execution system, device or device use or
Person's program in connection.The program code for including on computer-readable medium can transmit with any suitable medium,
Including but not limited to: electric wire, optical cable, RF (radio frequency) etc. or above-mentioned any appropriate combination.
Above-mentioned computer-readable medium can be included in above-mentioned electronic equipment；It is also possible to individualism, and not
It is fitted into the electronic equipment.Above-mentioned computer-readable medium carries one or more program, when said one or more
When a program is executed by the electronic equipment, so that the electronic equipment: obtaining audio-frequency information to be converted；Audio-frequency information is converted into
Text information；Text information is subjected to word cutting, obtains word sequence；For the word in word sequence, connected by word trained in advance
Word that probabilistic model obtains connection probability tables inquire the word and connected between next word of the word probability and the word with it is all kinds of
The connection probability of punctuate, and the linking objective based on the connection determine the probability word inquired；By word each in word sequence and phase
The linking objective answered connects article of the generation with punctuate and is exported.
The behaviour for executing embodiment of the disclosure can be write with one or more programming languages or combinations thereof
The computer program code of work, described program design language include object oriented program language-such as Java,
Smalltalk, C++ further include conventional procedural programming language-such as " C " language or similar program design language
Speech.Program code can be executed fully on the user computer, partly be executed on the user computer, as an independence
Software package execute, part on the user computer part execute on the remote computer or completely in remote computer or
It is executed on server.In situations involving remote computers, remote computer can pass through the network of any kind --- packet
It includes local area network (LAN) or wide area network (WAN)-is connected to subscriber computer, or, it may be connected to outer computer (such as benefit
It is connected with ISP by internet).
Flow chart and block diagram in attached drawing are illustrated according to the system of the various embodiments of the disclosure, method and computer journey
The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation
A part of one module, program segment or code of table, a part of the module, program segment or code include one or more use
The executable instruction of the logic function as defined in realizing.It should also be noted that in some implementations as replacements, being marked in box
The function of note can also occur in a different order than that indicated in the drawings.For example, two boxes succeedingly indicated are actually
It can be basically executed in parallel, they can also be executed in the opposite order sometimes, and this depends on the function involved.Also it to infuse
Meaning, the combination of each box in block diagram and or flow chart and the box in block diagram and or flow chart can be with holding
The dedicated hardware based system of functions or operations as defined in row is realized, or can use specialized hardware and computer instruction
Combination realize.
Being described in unit involved in embodiment of the disclosure can be realized by way of software, can also be passed through
The mode of hardware is realized.Described unit also can be set in the processor, for example, can be described as: a kind of processor
Including acquiring unit, converting unit, word cutting unit, judging unit and connection unit.Wherein, the title of these units is in certain feelings
The restriction to the unit itself is not constituted under condition, for example, acquiring unit is also described as " obtaining audio letter to be converted
The unit of breath ".
Above description is only the preferred embodiment of the disclosure and the explanation to institute's application technology principle.Those skilled in the art
Member is it should be appreciated that invention scope involved in the disclosure, however it is not limited to technology made of the specific combination of above-mentioned technical characteristic
Scheme, while should also cover in the case where not departing from the inventive concept, it is carried out by above-mentioned technical characteristic or its equivalent feature
Any combination and the other technical solutions formed.Such as features described above has similar function with (but being not limited to) disclosed in the disclosure
Can technical characteristic replaced mutually and the technical solution that is formed.
Claims (14)
1. a kind of method for output information, comprising:
Obtain audio-frequency information to be converted；
The audio-frequency information is converted into text information；
The text information is subjected to word cutting, obtains word sequence；
For the word in the word sequence, the word connection probability tables inquiry that probabilistic model obtains is connected by word trained in advance and is somebody's turn to do
Word and the connection probability that probability and the word and all kinds of punctuates are connected between next word of the word, and based on the company inquired
Connect the linking objective of the determine the probability word；
Word each in the word sequence is connected article of the generation with punctuate with corresponding linking objective to export.
2. according to the method described in claim 1, wherein, institute's predicate connection probability tables are obtained through the following steps:
Obtain training sample set, wherein training sample includes the sentence containing punctuate；
Using the sentence of the training sample in the training sample set as the input of LSTM model, training obtains word connection probability
Model；
Connected according to institute's predicate obtained in the pilot process of probabilistic model training probability between each word and word and each word with
Probability between each punctuate generates word and connects probability tables.
3. according to the method described in claim 2, wherein, the acquisition training sample set, comprising:
Sample article is obtained, the sample article is subjected to cutting by the granularity of one big sentence and obtains sample sentence set, wherein is big
Sentence refers to the sentence to end up with fullstop, question mark or exclamation mark；
For the sample sentence in the sample sentence set, term vector is generated after which is carried out word cutting as training sample.
4. according to the method described in claim 1, wherein, the method also includes:
The article is divided at least one paragraph.
5. method described in one of -4 according to claim 1, wherein the method also includes:
Determine the theme and entity of the article；
It obtains and the theme of the article and the image of Entities Matching；
Graph text information is generated according to described image and the article.
6. according to the method described in claim 5, wherein, the method also includes:
The graph text information is subjected to typesetting optimization.
7. a kind of device for output information, comprising:
Acquiring unit is configured to obtain audio-frequency information to be converted；
Converting unit is configured to the audio-frequency information being converted into text information；
Word cutting unit is configured to the text information carrying out word cutting, obtains word sequence；
Judging unit, is configured to for the word in the word sequence, connects what probabilistic model obtained by word trained in advance
Word connection probability tables inquire the word and connect the connection probability of probability and the word and all kinds of punctuates between next word of the word,
And the linking objective based on the connection determine the probability word inquired；
Connection unit is configured to word each in the word sequence connecting text of the generation with punctuate with corresponding linking objective
Zhang Jinhang output.
8. device according to claim 7, wherein described device further includes training unit, is configured to:
Obtain training sample set, wherein training sample includes the sentence containing punctuate；
Using the sentence of the training sample in the training sample set as the input of LSTM model, training obtains word connection probability
Model；
Connected according to institute's predicate obtained in the pilot process of probabilistic model training probability between each word and word and each word with
Probability between each punctuate generates word and connects probability tables.
9. device according to claim 8, wherein the training unit is further configured to:
Sample article is obtained, the sample article is subjected to cutting by the granularity of one big sentence and obtains sample sentence set, wherein is big
Sentence refers to the sentence to end up with fullstop, question mark or exclamation mark；
For the sample sentence in the sample sentence set, term vector is generated after which is carried out word cutting as training sample.
10. device according to claim 7, wherein described device further includes segmenting unit, is configured to:
The article is divided at least one paragraph.
11. the device according to one of claim 7-10, wherein described device further includes figure unit, is configured to:
Determine the theme and entity of the article；
It obtains and the theme of the article and the image of Entities Matching；
Graph text information is generated according to described image and the article.
12. device according to claim 11, wherein described device further includes typesetting unit, is configured to:
The graph text information is subjected to typesetting optimization.
13. a kind of electronic equipment, comprising:
One or more processors；
Storage device is stored thereon with one or more programs,
When one or more of programs are executed by one or more of processors, so that one or more of processors are real
Now such as method as claimed in any one of claims 1 to 6.
14. a kind of computer-readable medium, is stored thereon with computer program, wherein real when described program is executed by processor
Now such as method as claimed in any one of claims 1 to 6.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910552619.6A CN110245334B (en) | 2019-06-25 | 2019-06-25 | Method and device for outputting information |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910552619.6A CN110245334B (en) | 2019-06-25 | 2019-06-25 | Method and device for outputting information |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN110245334A true CN110245334A (en) | 2019-09-17 |
| CN110245334B CN110245334B (en) | 2023-06-16 |
Family
ID=67889231
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910552619.6A Active CN110245334B (en) | 2019-06-25 | 2019-06-25 | Method and device for outputting information |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN110245334B (en) |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111078831A (en) * | 2019-11-06 | 2020-04-28 | 广州荔支网络技术有限公司 | Optimization method for converting audio content into text in text reading |
| CN113297824A (en) * | 2021-05-11 | 2021-08-24 | 北京字跳网络技术有限公司 | Text display method and device, electronic equipment and storage medium |
| WO2022143058A1 (en) * | 2021-01-04 | 2022-07-07 | 北京有竹居网络技术有限公司 | Voice recognition method and apparatus, storage medium, and electronic device |
Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070203887A1 (en) * | 2004-06-30 | 2007-08-30 | Misha Dynin | Methods and systems for endorsing search results |
| US20130132401A1 (en) * | 2011-11-17 | 2013-05-23 | Yahoo! Inc. | Related news articles |
| CN106951413A (en) * | 2017-03-24 | 2017-07-14 | 北京百度网讯科技有限公司 | Segmenting method and device based on artificial intelligence |
| CN107221330A (en) * | 2017-05-26 | 2017-09-29 | 北京搜狗科技发展有限公司 | Punctuate adding method and device, the device added for punctuate |
| CN107679039A (en) * | 2017-10-17 | 2018-02-09 | 北京百度网讯科技有限公司 | The method and apparatus being intended to for determining sentence |
| CN107767870A (en) * | 2017-09-29 | 2018-03-06 | 百度在线网络技术（北京）有限公司 | Adding method, device and the computer equipment of punctuation mark |
| CN107944032A (en) * | 2017-12-13 | 2018-04-20 | 北京百度网讯科技有限公司 | Method and apparatus for generating information |
| CN108920611A (en) * | 2018-06-28 | 2018-11-30 | 北京百度网讯科技有限公司 | article generation method, device, equipment and storage medium |
| CN108932220A (en) * | 2018-06-29 | 2018-12-04 | 北京百度网讯科技有限公司 | article generation method and device |
| CN108959556A (en) * | 2018-06-29 | 2018-12-07 | 北京百度网讯科技有限公司 | Entity answering method, device and terminal neural network based |
| CN109213997A (en) * | 2018-08-16 | 2019-01-15 | 昆明理工大学 | A kind of Chinese word cutting method based on two-way length memory network model in short-term |
| CN109241330A (en) * | 2018-08-20 | 2019-01-18 | 北京百度网讯科技有限公司 | The method, apparatus, equipment and medium of key phrase in audio for identification |
| CN109657041A (en) * | 2018-12-04 | 2019-04-19 | 南京理工大学 | The problem of based on deep learning automatic generation method |
-
2019
- 2019-06-25 CN CN201910552619.6A patent/CN110245334B/en active Active
Patent Citations (13)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20070203887A1 (en) * | 2004-06-30 | 2007-08-30 | Misha Dynin | Methods and systems for endorsing search results |
| US20130132401A1 (en) * | 2011-11-17 | 2013-05-23 | Yahoo! Inc. | Related news articles |
| CN106951413A (en) * | 2017-03-24 | 2017-07-14 | 北京百度网讯科技有限公司 | Segmenting method and device based on artificial intelligence |
| CN107221330A (en) * | 2017-05-26 | 2017-09-29 | 北京搜狗科技发展有限公司 | Punctuate adding method and device, the device added for punctuate |
| CN107767870A (en) * | 2017-09-29 | 2018-03-06 | 百度在线网络技术（北京）有限公司 | Adding method, device and the computer equipment of punctuation mark |
| CN107679039A (en) * | 2017-10-17 | 2018-02-09 | 北京百度网讯科技有限公司 | The method and apparatus being intended to for determining sentence |
| CN107944032A (en) * | 2017-12-13 | 2018-04-20 | 北京百度网讯科技有限公司 | Method and apparatus for generating information |
| CN108920611A (en) * | 2018-06-28 | 2018-11-30 | 北京百度网讯科技有限公司 | article generation method, device, equipment and storage medium |
| CN108932220A (en) * | 2018-06-29 | 2018-12-04 | 北京百度网讯科技有限公司 | article generation method and device |
| CN108959556A (en) * | 2018-06-29 | 2018-12-07 | 北京百度网讯科技有限公司 | Entity answering method, device and terminal neural network based |
| CN109213997A (en) * | 2018-08-16 | 2019-01-15 | 昆明理工大学 | A kind of Chinese word cutting method based on two-way length memory network model in short-term |
| CN109241330A (en) * | 2018-08-20 | 2019-01-18 | 北京百度网讯科技有限公司 | The method, apparatus, equipment and medium of key phrase in audio for identification |
| CN109657041A (en) * | 2018-12-04 | 2019-04-19 | 南京理工大学 | The problem of based on deep learning automatic generation method |
Non-Patent Citations (2)
| Title |
|---|
| 侯加英: ""汉越双语新闻话题发现研究"", 《中国优秀硕士学位论文全文数据库信息科技辑》, pages 138 - 1937 * |
| 江跃华等: ""融合词汇特征的生成式摘要模型"", 《河北科技大学学报》, pages 152 - 158 * |
Cited By (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111078831A (en) * | 2019-11-06 | 2020-04-28 | 广州荔支网络技术有限公司 | Optimization method for converting audio content into text in text reading |
| CN111078831B (en) * | 2019-11-06 | 2023-05-30 | 广州荔支网络技术有限公司 | Optimizing method for converting text from text reading audio content |
| WO2022143058A1 (en) * | 2021-01-04 | 2022-07-07 | 北京有竹居网络技术有限公司 | Voice recognition method and apparatus, storage medium, and electronic device |
| CN113297824A (en) * | 2021-05-11 | 2021-08-24 | 北京字跳网络技术有限公司 | Text display method and device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN110245334B (en) | 2023-06-16 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108022586B (en) | Method and apparatus for controlling the page | |
| CN110298906B (en) | Method and device for generating information | |
| US10956480B2 (en) | System and method for generating dialogue graphs | |
| CN110321544B (en) | Method and device for generating information | |
| CN106227792B (en) | Method and apparatus for pushed information | |
| US11990124B2 (en) | Language model prediction of API call invocations and verbal responses | |
| CN109643540A (en) | System and method for artificial intelligent voice evolution | |
| CN110245334A (en) | Method and apparatus for output information | |
| CN109829164A (en) | Method and apparatus for generating text | |
| WO2021063089A1 (en) | Rule matching method, rule matching apparatus, storage medium and electronic device | |
| WO2023005968A1 (en) | Text category recognition method and apparatus, and electronic device and storage medium | |
| CN111753551A (en) | Information generation method and device based on word vector generation model | |
| CN111582360A (en) | Method, apparatus, device and medium for labeling data | |
| US20210004603A1 (en) | Method and apparatus for determining (raw) video materials for news | |
| CN112364653A (en) | Text analysis method, apparatus, server and medium for speech synthesis | |
| WO2024099171A1 (en) | Video generation method and apparatus | |
| CN112182255A (en) | Method and apparatus for storing media files and for retrieving media files | |
| CN110457325B (en) | Method and apparatus for outputting information | |
| CN111555960A (en) | Method for generating information | |
| CN111008213A (en) | Method and apparatus for generating language conversion model | |
| CN109710939A (en) | Method and apparatus for determining theme | |
| CN112633004A (en) | Text punctuation deletion method and device, electronic equipment and storage medium | |
| CN109241296A (en) | Method and apparatus for generating information | |
| CN109947526A (en) | Method and apparatus for output information | |
| CN112652329B (en) | Text realignment method and device, electronic equipment and storage medium |
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 |