Online Journal of Applied Knowledge Management

A Publication of the International Institute for Applied Knowledge Management

Volume 3, Issue 1, 2015

The role of semantics in searching for information

on the Web
Ilona Pawełoszek, Czestochowa University of Technology, ipaweloszek@zim.pcz.pl

Jędrzej Wieczorkowski, Warsaw School of Economics, jedrzej.wieczorkowski@sgh.waw.pl

Abstract
As the amount of data in the World Wide Web grows, the Internet becomes the biggest and often the primary
information resource for many individuals and organizations. To make good use of the data, it is essential to provide
effective and intelligent search capabilities. The users more and more often require direct and unambiguous answers,
which are often not explicitly present in any document. The Web users also are not willing to learn complex query
languages and need interfaces that will be easy to use and as close as possible to natural language. The aim of this
paper is to illustrate the advantages that semantics brings to the users of contemporary Web search engines. The
concepts of semantic Web and semantic search engines have been described along with some issues related to their
development and usage, such as linked data, semantic query interfaces and the ways of publishing semantic data. The
authors also explore the semantic features of contemporary search engines and indicate their future directions of
development.

Keywords: Semantic Web, search engine, linked data, semantic query interfaces, semantic data publishing

Introduction

The World Wide Web has come a long way since its public introduction in April 1993. From that
moment the WWW technology has been available for anyone to use on a royalty-free basis
(History of the Web 2015). The number of Web pages sprung out rapidly and it is constantly
growing. Also the internet services increase in number and sophistication. From once static pages,
Internet sites have evolved and specialized in offering interactive and device-customized content.
Despite the rapid development of the Web technologies, still the main purpose of the Web is to be
exploited by human users, therefore the techniques of building websites mostly have been oriented
on visual presentation of information and data.

More and more knowledge workers use the Web as a primary source of information in their daily
tasks. A very important issue is that information in the Web documents often lacks structure and
is many times rather complicated to read and understand. Data are often presented in tables
interspersed with text in lengthy articles, whereas the user needs to interpret and extract the
meaningful pieces of data. The internauts search for information, browse the content, download
and make use of it in off-line and on-line applications. The use of search engines is the most
popular way of accessing information on the Web. Since the very inception, the search engines
have been single points of access to different kinds of Web content. The semantic Web
technologies, which help to organize information in a meaningful way that enables reasoning as
well by computers as by human users are promising and potentially important tools to make
extensive use of Web resources.

102
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 3, Issue 1, 2015

The aim of this paper is to illustrate the advantages that semantics bring to the users of
contemporary Web search engines. We also discuss some difficulties in applying the semantic
approach. Section 1 introduces the reader to the idea of the semantic Web and semantic search.
The discussion on semantic features already implemented in contemporary search engines on the
example of google.pl has been presented in the section 2. In the section 3 the authors attempt to
specify the categories of search- related problems that can be solved by exploiting semantics.
Semantic Web as an emerging technology platform also creates challenges for companies trying
to build their online presence and position their websites in the search engine results. The ways in
which to present the company’s data on the Semantic Web are also described in section 3. Finally,
we conclude our considerations and follow that up with the future work that can be done to enhance
semantic capabilities of today’s Web applications.

The semantic Web and semantic search – motivation and definition

In most Web search engines the list of retrieved documents is displayed after typing the keywords
specifying the user’s information-seeking goals. In the present scenario well-grounded keyword
searching seems outdated and not responding to users’ needs. The present generation search
engines present a large amount of search results to the user in response to the query. Depending
on the keywords and the way the query is posed some of the results are more or less irrelevant, and
the user has an ordeal in sifting through the result sets to harvest some information of his interest
(Mittal, Singh & Sachdeva 2011). Sometimes the answer to the user’s query is not directly present
on any Web document. In such a case it can only be inferred by analyzing and extracting data from
many resources. This task, unfortunately so far must be done manually. The limitations of
contemporary search engines have turned the researchers to seek for new alternatives which led to
the emergence of the semantic approach to organize and search for information resources. Since
the idea of semantic Web had been introduced in 2001 by Tim Berners-Lee (Berners-Lee, Hendler,
& Lassila 2001). Many interesting and progressive technologies have emerged to pursue the vision
of a Web that will not only simply contain data, but semantic information that is machine-
processable in a meaningful way (Stoermer 2006). Semantics is the study of the meaning of words,
phrases and sentences that are associated with the real-world objects and concepts. Semantic data
structures are inspired by human natural language constructs in a way that they present knowledge
as simple indicative sentences having subject, predicate and object.

There are many advanced solutions that have been developed for many years, with the aim to help
computers to interpret human language, and communicate in a human-like ways, these are artificial
intelligence techniques such as natural language processing. However, they had not advanced to a
level that could be usable for the machine interpretation of Web documents authored by humans.
The original semantic approach envisioned by T. Berners-Lee and his coworkers was to keep the
solutions simple enough so they can be used by webmasters without experience in knowledge
engineering. Therefore the concept of the semantic web is based on a simple idea of annotating
Web documents with Extensible Markup Language (XML) markup which is in many aspects
similar to well-known and widely accepted Hypertext Markup Language (HTML) syntax.

103
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 3, Issue 1, 2015

The semantic annotations should correspond to some clearly defined and shared knowledge
structure, which is called ontology.

Ontology is a specification designed to categorize and help explain the relationships between
various concepts of in the given area of knowledge and research. The most widely accepted
definition of ontology in the context of knowledge sharing and information science was proposed
by T. Gruber (1995), he says that ontology is a specification of a conceptualization. The
interpretation of this definition can be that

• Ontology provides a common understanding of a particular domain, or field, of study, and

ensures a shared ground for those who study the domain.
• Ontology is useful for organizing concepts, information, and ideas, it helps to show the
relations between concepts.
• Ontology can be formalized which means that it can be read and understood by computer
applications.

The role of ontology in knowledge engineering is to enable construction of a domain model by

describing a set of concepts and relations between them (Jurczyk & Pawełoszek 2015).

The language for annotating documents is Resource Description Framework (RDF) and for
ontology definition RDF Schema (RDFS) or Web Ontology Language (OWL) are used. The
resources on the Web, i.e., the Web pages, are crawled and looked for annotations done on them,
if any. A semantic search engine uses ontologies to derive semantic associations among different
words and concepts. These annotations are then compared to that ontology with which they have
been tagged (Mittal, Singh & Sachdeva 2011).

According to a software company Cambridge Semantics Inc. headquartered in Boston,

Massachusetts: The Semantic Web is a set of technologies for representing, storing, and querying
information. Although these technologies can be used to store textual data, they typically are used
to store smaller bits of data (Cambridge Semantics 2014). Essentially, the semantic Web focuses
on pulling specific data (i.e., numbers, dates, locations) from multiple heterogeneous sources to
answer directly the users’ questions. Therefore the semantic Web is often described as a Web of
Data. In different words: it works more like a huge database than a collection of textual documents.
To make the Web of Data a reality, it is important to have the data on the Web available in a
standard format, reachable and manageable by Semantic Web tools (W3C 2015).

The data extracted from Web documents, describe some real-world object, a person or a concept
are generally referred to as entities. So called entity search fully reflects the idea of the semantic
web, which is searching for the entities by asking about their attributes instead of searching for
text describing the entity. T. Doszkocs (2012) explains the idea of semantic search as “a search or
a question or an action that produces meaningful results, even when the retrieved items contain
none of the query terms, or the search involves no query text at all”. A good example is a question
about a person’s age. The user can specify a query as follows: “how old is person X”. The answer

104
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 3, Issue 1, 2015

may not be directly written in any document, although it can be deduced from the data about the
person’s X Birthdate. The semantic search requires reasoning capabilities and knowledge base
containing ontologies.

Semantics in modern search engines

When studying the publications discussing the need for semantic Web and semantic search it is
easy to find the opinions that using search engines is difficult, the search results are mostly
irrelevant and the user has to browse many documents to find what he really needs. The common
opinion is that the search engines are “unintelligent” and cannot understand the context of the
user’s query. The current generation of search engines is severely limited in its understanding of
the user’s intent and the Web content and consequently in matching the needs for information with
the resources on the Web (Mika 2008). However, the study of the most popular search engines
reveals that the aforementioned facts are not always so obvious. In the recent years the search
engines are increasingly moving towards semantically enhanced results for user queries.

Contemporary search engines allow the user to specify the query and obtain better/more relevant
results by offering many semantics-related options narrowing the results or changing their order.
It is possible to specify a search query for text, images, videos and services such as maps, news,
books etc. Users can also choose the desired language or domain (i.e. pcz.pl), however the latter
requires knowledge how to use advanced options. The search engines offer many options and the
possibility to use search operators to make the user’s query more specific and to obtain better (more
relevant) results. Moreover the advanced search options are quite easy to use (assuming that the
user is aware of their existence).

Some of the contemporary search engines are context-sensitive. This means they already exploit
some semantics. Very often it is easy to observe that search results are relevant to the user's location
– for example, when the user localized in Katowice will type the query “restaurant” the first search
results will probably be the links to the websites of the restaurants nearby the user’s location.

The next “intelligent” feature of modern search engines is the awareness of synonyms and spelling
mistakes. For example, if we type a query “information searching” (without quotes) the Google
search engine also highlights the word “seeking” which is synonymous to “searching” (figure 1.).

105
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 3, Issue 1, 2015

Figure 2. Example of the awareness of synonyms in Google search results

Source: Chrome browser and http://google.pl

While querying Google i.e. for the word “fox” it will return many results, but their sequence will
be dependent on the language the user speaks. In case of Polish language the first search results
will pertain to Fox TV channel, Fox Broadcasting Company. Although there is some inconsistence
between text and image search results. The image search will return the pictures of fox – the
animal. In many cases typing the foreign word or phrase will result in links to dictionaries and
translators. The search engines can also serve as measurement and currency converters.

One of the significant changes in the ways of interaction with search interfaces is the conversational
search presented by Google in its new Chrome browser. Conversational search is a new kind of
philosophy for human-computer interaction. The principle behind it is that the user can speak a
sentence into a device, and that device can respond with a full sentence (Techopedia 2015). The
idea of conversational search is to make the human-computer interaction more natural and
convenient for people. The feature is especially meant for mobile devices, while voice interface is
more convenient for users while being on the go.

Conversational search exploits the natural language processing and speech recognition (Google's
Voice Search) to formulate the query string. The generated query string is then analyzed to extract
keywords. The keywords can be interpreted by semantic search mechanisms, which try to
formulate the semantic query, usually in SPARQL (W3C 2008) language. Then, if possible, the
full-sentence answer for the query is generated. The possibility of performing semantic search and
answering with the full sentence is dependent on many conditions, the most important are:
availability of semantic data sources for a given domain, the scope of semantic data sources and

106
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 3, Issue 1, 2015

underlying ontologies, and the accurateness of Automatic Speech Recognition (ASR) techniques.
As the research shows the ASR engines may provide satisfying transcription (Silber-Varod, Geri
2014).

Another interesting concept (although not yet fully implemented) is called “previous query” – the
idea is that the conversational search remembers the user’s question and takes it into consideration
giving the answers for the consecutive searches. For example, if someone searches for [Poland]
and then [president] after that, the results should be altered to take the previous query into account.
To some degree, it will be as if the second query was for [president of Poland].

Semantic search problems and solutions

Semantic search problems can be regarded from the point of view of search engine developer, user
and webmaster. From the point of view of application developer the basic practical problem in
building semantic applications is how to extract entities from the text. If the entities are extracted
the next step is to interpret the values of their attributes, compare them to the domain ontologies,
infer the answer and display it to the user.

Extracting data from the webpage is not a trivial task. There are two approaches to building
semantic applications. The first one assumes that the text contains markup explicitly identifying
the entities and their attributes. In this case the search engine consumes structured data. The second
approach (the implicit one) assumes that the search engine deals with unstructured text, therefore
the advanced algorithms (stochastic, NLP) must be used to derive or infer the data.

Both the approaches have some advantages and pitfalls. The first - explicit approach requires from
the webmasters the efforts of annotating webpages. The task of annotating is not so straightforward
because it must be done manually, some programming environments do not have functions to
support the user in adding annotations in the specific markup standard. The second issue is the
decision about which ontology to choose as the reference model for annotations. The good thing
about explicit approach is the fact that the annotations made by human are usually correct in the
logical sense. Although the search engine based on this approach can only operate on the set of
documents which had been annotated.

The second - implicit approach does not require annotated documents, therefore theoretically the
search engine can work on any text documents found on the Web. The disadvantage of this
approach is that it strongly depends on artificial intelligence algorithms and text mining which are
not perfect and usually are restricted to some specific domain. Sometimes the context can be
misunderstood and the entities or their attributes will not be properly identified. The implicit
approach is often used in the cases such as: finding documents on a particular subject or domain,
determining the sentiment of the documents (positive or negative context).

The third consideration of the semantic application developer is which ontologies to choose or if
there are no suitable ontologies he may decide to build it on his own. The developers often look

107
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 3, Issue 1, 2015

for appropriate ontologies that can be integrated into their systems, rather than develop new
ontologies from scratch. The choice of ontology may, have a major impact on the performance of
the semantic application, including the quality of the results (Tan & Lambrix 2009). In case of
building large ontology it is possible to integrate several existing ontologies describing portions of
the large domain. One can also reuse a general ontology or taxonomy, such as the UNSPSC (United
Nations Standard Products and Services Code) or PKWiU (Polish Classification of Goods and
Services), and extend it to describe the domain of interest (Noy & McGuiness 2001).

The success of semantic applications strongly depends on its query interface. Users are not willing
to get used to complicated kinds interaction. Therefore semantic search should not be much
different than keyword search that the users are used to. Otherwise the search engine will not be
used. On one hand the semantic Web query languages e.g. SPARQL (Prud'hommeaux & Seaborne
2008), SeRQL (Broekstra & Kampman 2004), RDQL (Seaborne 2004), are difficult to use, but on
the other hand the strongly formalized graph-based approach represented by those languages
allows for very precise query definition and addresses complex information needs. Among the
aforementioned query languages SPARQL is the one that has been recognized as the de facto
standard for the Semantic Web. The most user-friendly solution for query definition is undouble
natural language, however it has unavoidably lower accuracy, as compared to systems with graph-
based querying interfaces, which, in turn, are usually still too difficult for regular users (Styperek,
Ciesielczyk & Szwabe 2014). Using semantic query languages poses two main challenges for the
users, first of all they must be familiar with the language syntax, and the second important
requirement is the knowledge of underlying ontologies. There are many propositions of semantic
search engine interfaces. They usually present the tradeoff between the easiness of use and the
semantic capabilities.

The architecture presented in (Wang et al 2008) proposes a solution to translate keyword queries
to formal queries. The Authors leverage terms extracted from Wikipedia to enrich literals described
in the original RDF data. This way, users need not use keywords that exactly match the RDF data.

Another interesting proposition (Bäurle 2011) combines and extends established components of
other search user interfaces - namely keyword search, facets, proposals and breadcrumbs. The user
interface provides only a single input field as it is also common for current keyword search
applications. When something is typed, it automatically shows different proposals for words,
relations, entities, or semantic classes that are used to build the search query.

From the perspective of the webmaster a crucial problem is how join the semantic Web, this means
publishing the data for more effective discovery, automation, integration, and reuse. It is especially
important for internet marketers and the SEO (Search Engine Optimization) professionals. At
present the most popular search engine is Google, and it also is making great efforts to develop its
semantic capabilities. It is more and more often that after posing a question in the natural language
one can get the exact answer. For example a question about the age of Polish president Bronislaw
Komorowski results in very precise answer displayed along with some other information about the
person (Figure 2). Moreover the semantic features are dependent on language version of the user’s

108
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 3, Issue 1, 2015

operating system. The example on Figure 2 also illustrates another important concept strongly
related to semantic Web – the Linked Data, which is a collection of interrelated datasets on the
Web. The query to the semantic search engine displays desired entity and its attributes, but also
proposes some other entities somehow related to the one the user is asking about. In the case of
the query for Polish president, the search engine displays also two important Polish politicians,
and the president’s wife. The entities are connected with some mutual relationships of different
kinds. The concept of Linked Data is about using the Web to connect related data (Linked Data
2015). The Linked data concept emphasizes that not only the Semantic Web does need access to
data, but relationships among data should be made available.

Figure 2. The display of semantic search results

Source: Chrome browser and http://google.pl

The entity (a person or a business unit) must comply with some requirements to be found by
semantic search engine. The requirements for being displayed in semantic search results are
different than in case of traditional SEO. There are three basic methods to serve linked data on the
Web:

• Google Knowledge Graph (),

• publishing RDF files,
• labeling HTML content with microformats, microdata or RDFa.

Google Knowledge Graph is the secret behind the entity-based search and the straightforward
immediate answers to the user’s question posed to Google Search engine. The Google Knowledge
Graph, includes information about the entities along with their relationships. The Knowledge
Graph takes information from trusted data sources such as Freebase (2015), Wikipedia (2015) and
Google+ (2015) account, so it is recommended to join the communities and fill one’s profile.

109
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 3, Issue 1, 2015

However recently Google has been developing a new generation base for semantic search which
is called Knowledge Vault (Dong et. al. 2014) . The difference between Google's existing
Knowledge Graph and the Knowledge Vault is the way that facts are accumulated. The Knowledge
Graph acquires information from trusted sources. The Knowledge Vault is an accumulation of
facts from Web content (obtained via analysis of text, tabular data, page structure, and human
annotations) with prior knowledge derived from existing knowledge repositories. The major
difficulty is that, by its very nature, the Semantic Web is a large, uncensored system to which
anyone may contribute. This raises the question of how much credence to give each source.

The next generation solution employs supervised machine learning methods for fusing distinct
information sources and probabilistic inference methods that compute probabilities of fact
correctness (Dong et al 2014).

Publishing RDF files on the Web requires the webmaster to provide some external URLs pointing
to them, so that crawlers can discover the new added data. The RDF data source can be added to
the ESW Wiki datasets list, also links can be added from one’s Friend of a Friend (FOAF) profile.

The lightway semantisation for non-technical users can be achieved by adding so called
microformats to the HTML websites. Microformats (2015) are XHTML tags for marking up
people, organizations, events, locations, blog posts, products, reviews, resumes, recipes etc. A data
with microformats can be consumed and used by search engines, browsers, and other sites.
Examples of well-established microformats are:

• hCalendar – events,
• hCard - people, organizations, contacts,
• rel-license - licensed content,
• rel-nofollow - links in untrusted 3rd party content,
• rel-tag - tag posts and pages by subject,
• XFN - social relationships and rel-me links among profiles for the same person, XMDP
- define a microformat vocabulary / profile, XOXO – outlines.

Microdata is an HTML specification to label content to describe a specific type of information, for
example, reviews, persons, or events and their properties. For example, an event has the properties,
such as: venue, starting time, name, and category.

RDFa is an extension to HTML5, designed for labeling content. It uses simple attributes in
XHTML tags (often <span> or <div>) to assign brief and descriptive names to entities and
properties. RDFa annotation tools are added to popular web editing tools For example, a plugin
AKSW (2015) added for TinyMCE (2015) editor which is a platform independent web based
Javascript HTML WYSIWYG editor often implemented in Open Source Content Management
Systems platforms.

110
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 3, Issue 1, 2015

All the aforementioned techniques allow the webmasters to markup things like People, Places,
Events, Recipes and Reviews. Search Engines and Web Services use this markup to generate better
search listings.

Conclusions

Although there are many initiatives the Semantic Web has not reached its full potential yet. The
real added value from the semantic Web has not yet been achieved due to many problems that must
be faced before the intelligent Internet can be exploited as a reliable source of knowledge that can
be processed in a fully automatic way. On the other hand the problems become a challenge for
development of new business models (for example, providers of ontologies, trust services,
semantic cloud services, semantic search interfaces and many more).

A very interesting future concept is the semantic personal data locker (Lockerproject 2015), which
is a kind of the user’s digital profile. The personal data locker contains the user’s detailed data
(such as preferences, professional and private schedules, memos, etc.) which can be used as the
context parameters for the semantic applications. It would be for example possible to integrate the
data from the user’s shopping list with the information from the Web about sales and promotions
and thus provide fully relevant and interesting search results. The idea of personal data locker
assumes storing all the data in the cloud. Although the data are to be under total control of the user
the concept still gives rise to many privacy concerns.

More and more companies and public institutions are discovering the power of the Internet in
reaching their customers and expanding their marketing influence into previously unreachable
localizations. While the Web increases in volume and heterogeneity, it becomes increasingly
important resource of information and the area of business activity. Even the smallest organizations
are quickly realizing the need for internet presence which is undeniable because individuals and
businesses today predominantly make buying decisions based on what they discover online. The
Semantic Web is often envisioned as the future of the Internet, therefore; it seems essential for the
companies to keep an eye on the development of new technologies and adopt them to gain
competitive advantage on the electronic market.

References

Bäurle, F. (2011). Master's thesis, University of Freiburg, Retrieved January 19, 2015 from
https://ad.informatik.uni-freiburg.de/files/ui_semantic_fulltext_search
Berners-Lee, T., Hendler, J. & Lassila, O. (2001). The Semantic Web, Scientific American, May
2001, (pp. 29-37)
Broekstra, J. & Kampman, A. (2004). SeRQL: An RDF Query and Transformation Language,3rd
International Semantic Web Conference. Japan.

Cambridge Semantics (2014, November 17). Semantic Search and the Semantic Web, Retrieved

111
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 3, Issue 1, 2015

January 19, 2015 from http://www.cambridgesemantics.com/semanticuniversity/semantic-

search-and-semantic-web
Dong, X.L., Murphy, K., Gabrilovich, E., Heitz, G., Horn, W., Lao, N., Strohmann,T., Sun S.&
Zhang W. (2014). Knowledge Vault: A Web-scale approach to probabilistic knowledge
fusion, Retrieved January 19, 2015 from
http://www.cs.ubc.ca/~murphyk/Papers/kvkdd14.pdf
Doszkocs, T. (2012). Semantics for universal search and discovery, Retrieved January 19, 2015,
from http://www.slideshare.net/jendre/semantics-for-universal-search-and-discover7
Gruber, T., 1995, Toward Principles for the Design of Ontologies Used for Knowledge Sharing.
International Journal Human-Computer Studies Vol. 43, Issues 5-6, November, p.907-
928, Retrieved March 29, 2015 from: http://tomgruber.org/writing/onto-design.pdf
History of the Web (n.d.). Retrieved January 19, 2015 from:
http://webfoundation.org/about/vision/history-of-the-web/
Jurczyk, M. & Pawełoszek, I., The Concept of Semantic System for Supporting Planning of
Innovation Processes, Polish Journal of Management Studies vol. 11/2015 (accepted for
publication)
Linked Data (2015). Retrieved January 19, 2015 from http://linkeddata.org/
Lockerproject (2015). Retrieved January 19, 2015 from http://lockerproject.org/
Microformats (2015). Retrieved January 19, 2015 from http://microformats.org/
Mika, P. (2008). Microsearch: An Interface for Semantic Search, Proceedings of the Workshop
on Semantic Search (SemSearch 2008) at the 5th European Semantic Web Conference
ESWC 2008 volume 334 of CEUR Workshop Proceedings.
Mittal, H., Singh, J. & Sachdcva, J. (2011). ARAGOG Semantic Search Engine: Working,
Implementation and Comparison with Keywoixi-Based Search Engines, Pred, A., Dietz,
J.L.G., Liu, K., Filipc, J. (eds.) IC3K 2009. CCIS. vol. 128. (pp. 177-186). Springer,
Heidelberg.
Noy, N.F. & McGuiness, D.L. (2001). Ontology Development 101: A Guide to Creating Your
First Ontology. Technical Stanford Knowledge Systems Laboratory Technical Report
KSL-01-05 and Stanford Medical Informatics Technical Report SMI-2001-0880 Retreived
January 19, 2015 from
http://protege.stanford.edu/publications/ontology_development/ontology101-
noymcguinness.html
Prud'hommeaux, E. & Seaborne, A. (2008). SPARQL Query Language for RDF. Retrieved
January 19, 2015 from http://www.w3.org/TR/rdf-sparql-query/
Seaborne, A. (2004). RDQL - a query language for RDF. W3C member submission, Retrieved
January 19, 2015 from http://www.w3.org/Submission/RDQL/
Silber-Varod, V. & Geri, N. (2014). Can automatic speech recognition be satisficing for
audio/video search? Keyword-focused analysis of Hebrew automatic and manual
transcription, Online Journal of Applied Knowledge Management A Publication of the
International Institute for Applied Knowledge Management Volume 2, Issue 1, 2014.
Retrieved March 25, 2015 from:
http://www.iiakm.org/ojakm/articles/2014/volume2_1/OJAKM_Volume2_1pp104121.pd
f

112
Online Journal of Applied Knowledge Management
A Publication of the International Institute for Applied Knowledge Management
Volume 3, Issue 1, 2015

Singhal, A (2012). Introducing the Knowledge Graph: Things, Not Strings. Google Official Blog.
Retrieved March 6, 2015 from:
http://googleblog.blogspot.com/2012/05/introducing-knowledge-graph-things-not.html.
Stoermer, H. (2006). Introducing Context into Semantic Web Knowledge Bases, Proceedings of
the CAISE*06 Doctoral Consortium, Luxembourg, 5th June 2006. Retrieved January 19,
2015, from http://ceur-ws.org/Vol-263/paper3.pdf
Styperek, A., Ciesielczyk, M. & Szwabe, A. (2014). Semantic search engine with an intuitive
user interface, Proceedings of the companion publication of the 23rd international
conference on World Wide Web companion, International World Wide Web Conferences
Steering Committee (pp. 383-384).
Tan, H. & Lambrix, P. (2009). Selecting an Ontology for Biomedical Text Mining, Proceedings
of the Workshop on BioNLP, pages 55–62, Boulder, Colorado, Association for
Computational Linguistics.
Techopedia (2015). Conversational Search, Retrieved January 19, 2015, from
http://www.techopedia.com/definition/29608/conversational-search.
W3C (2015). Linked Data, Retrieved January 19, 2015, from
http://www.w3.org/standards/semanticweb/data
Wang, H., Zhang, K., Liu, Q., Tran, T. & Yu, Y. (2008). Q2Seinantic: A lightweight keyword
interface to semantic search, Bechhofer, S., Hauswirlh, M., Hoffmann, J., Koubarakis, M.
(eds.) ESWC 2008. LNCS, vol. 5021, (pp. 584-598). Springer. Heidelberg.
Dong. X., Gabrilovich E., G Heitz, W Horn, N Lao, K Murphy, T Strohmann, Sun S. & Zhang,
W. (2014). Knowledge vault: a web-scale approach to probabilistic knowledge fusion. In
Proceedings of the 20th ACM SIGKDD international conference on Knowledge
discovery and data mining.

Web References
Freebase (2015) https://www.freebase.com/ Retrieved march 25, 2015
Wikipedia (2015) http://www.wikipedia.org/ Retrieved march 25, 2015
Google + (2015) https://plus.google.com/ Retrieved march 25, 2015
AKSW (2015) http://aksw.org/Projects/RDFaCE.html Retrieved march 25, 2015
TinyMCE (2015) http://www.tinymce.com/ Retrieved march 25, 2015
W3C (2008) SPARQL Query Language for RDF W3C Recommendation 15 January 2008
Retrieved march 25, 2015 from: http://www.w3.org/TR/rdf-sparql-query/

113