KR101697167B1 - An information retrieval method based-on semantic knowledge structures for learners using books' contents and index and the system thereof - Google Patents

Abstract

Semantic knowledge structure based information retrieval method that extracts table of contents and index list from publishing books belonging to domain in order to systematically provide domain knowledge and provides search result with index belonging to table of contents while providing domain knowledge in order of table of contents And a system, comprising: (a) a table of contents list and an index list of an issuer book, wherein the table of contents list includes a plurality of table entries arranged in order, the index list including a plurality of indexes; (b) retrieving the indexes belonging to the table of contents item from the index list; (c) searching the search sites for the searched indexes as keywords; (d) providing a search result of a search site as a search result of the searched index; And (e) providing the search results of the indexes belonging to the table of contents items in the order of the table of contents items in the table of contents list.
According to the semantic knowledge structure based information search method and system as described above, it is possible to acquire knowledge related to the overall and systematic knowledge structure of web users more easily in various rich web resources generated sporadically.

Description

[0001] The present invention relates to an information retrieval method and system based on a semantic knowledge structure using a table of contents and an index, and an information retrieval method and system based on semantic knowledge structure based on semantic knowledge structures for learners using books.

In order to systematically provide domain knowledge, the present invention provides a semantic knowledge structure structure that extracts a table of contents and a list of indexes from publishing books belonging to a domain, provides domain knowledge in the order of a table of contents and provides a search result with an index belonging to a table of contents Information retrieval method and system.

Bioinformatics Learners often combine books and Web searches to gain knowledge about bioinformatics. If so, why do bioinformatics learners need books while they are able to acquire knowledge through web search, and are they seeking additional knowledge through web search even with books?

Although it is possible to acquire knowledge through the web, the book needs a book to understand the overall contents and knowledge flow of bioinformatics. Even if there is a book, It is because of abundance.

A comprehensive overview of the entire knowledge flow of a book. According to Seeberg, the table of contents is a comprehensive and condensed representation of the author's thoughts. It is a systematic and general knowledge structure for the field or domain in question. (1).

However, the book has difficulties with limited or inexplicable description examples, inexplicable description difficulties, and quick updates.

In contrast, web content is generated by a large number of users, and thus it is possible to provide a rich and varied example of contents, difficulty in explanation, and quick creation and updating of contents.

Particularly, as the service supporting the participation of users along with the era of Web 2.0 is developed, the number of web contents generated by general users is rapidly increasing. In fact, for YouTube, YouTube users upload 24 hours of video every second [12] and uploaded over 14 billion videos by April 2010 [13] [4].

However, bioinformatics learners generally have to rely on keyword retrieval when searching the web. If learners do not have basic knowledge about bioinformatics, general keyword retrieval can be applied to web contents generated sporadically by many users. It is not easy to grasp the overall knowledge structure or to search for systematic information.

In addition, although the beginner learner (bioinformatics lecturer) who does not have background knowledge about bioinformatics and the learner (bioinformatics learner) who has more than a certain degree of knowledge have different knowledge about the field, There is a difficulty in retrieving Web contents.

Accordingly, the present invention proposes WeBook as a search service for providing meaningful web contents to bioinformatics learners in various web contents sporadically generated by a large number of users as a user participation service is developed.

We provide a method of searching based on the semantic knowledge structure to beginners of bioinformatics which is a beginner learner who does not have background knowledge about bioinformatics. By using general keyword search, general knowledge flow about bioinformatics or general knowledge I supplemented the search. Also, by providing a concept based search method to bioinformatics learners who have some knowledge about bioinformatics, we supplemented the general keyword search method which had to use the same search method regardless of learner 's knowledge level.

The WeBook proposed by the present invention aims at implementing a system for supporting semantic web search of bioinformatics learners by supplementing characteristics of book and web contents.

To this end, we created a conceptual network that shows semantic knowledge structures and related concepts together based on a systematic and overall knowledge structure, which is generated by bioinformatics experts. The mature semantic knowledge structure and the specific web service were created to enable the learner to search the web based on the semantic knowledge structure.

In addition, through the mashup with the Semantic Course Search Project, students are provided with information on bioinformatics related universities, lectures, lecture materials, and lecture plans.

The objectives of the present invention are as follows.

1. Provide a semantic knowledge structure as a retrieval method for bioinformatics beginners

2. Provide a conceptual network as a search method for bioinformatics

3. Semantic knowledge structure and mashup with specific site

4. Semantic Course Search Project mashup

The present invention is based on the book index and index of the book. Table of Contents explains that the author is a semantically linked general structure established for a particular field, and that people can grasp the relationship and flow between concepts through this table of contents [1] [2].

Egan. et al. [3] [5] proposed a table-based text browser, SuperBook (Figure 1), using the characteristics of this table of contents. Egan et al. Compared the search speeds of the SuperBook with general paper books. The search speed experiment was conducted to search the contents of the same book, and to compare the efficiency of the search method of the general paper book with the table of contents based browser SuperBook. Experiments by Egan et al. Have shown that SuperBook users have a higher search speed and accuracy, and in the open-book evaluation, SuperBook users have submitted higher-level essays. In addition, the subjects were more satisfied with the usability test of the SuperBook than the general paper books.

Although the SuperBook proposes the significance and applicability of a table of contents, it has limitations on the use of external content as it is an inventor of search efficiency of one book content.

Meriill defines the knowledge structure as "the relationship and structure between the constituent elements of knowledge" [6], and Wikipedia describes the knowledge structure as "a set of interrelationships between facts or knowledge in a particular field" 14].

In the present invention, a table of contents of a book is adopted to create a semantic knowledge structure for a specific field, and the following four factors, which are characteristics of a table of contents, are presented on the basis of this.

1. about a Domain (for your field)

2. by Domain Experts (by experts in your field)

3. with Domain Concepts (as applicable field)

4. Relationship and Structure

That is, the table of contents can be described as "the relationship and structure between the concepts represented by the concepts in the field by experts in the field about about a domain" And the table of contents is the same as the knowledge structure, which is a "set of mutual relations between facts and knowledge on a particular field".

In the present invention, the knowledge structure generated based on the table of contents is utilized as a semantic structure for searching and providing meaningful contents to learner in a massive amount of web contents generated sporadically.

Mash-up means that an artist or disc jockey mixes several songs into a single piece of music, which was originally used in pop music. The mashup, which means to mix, is being expanded to create new software, services, convergence applications, etc. by combining or combining information, services, and open API provided on the web [4] [10].

The housing map [16], which is one of the most representative examples of mashups, is a service that combines the Google Map API and real estate listing information [11], allowing users to easily search listing information of interest through the housing map 2).

A similar service, Real Estate Auction [17], is a service that provides real estate information registered on eBay [18] by mashing up the eBay API, the largest auction site in the US, and the Google Maps API (Figure 3).

The mashup, which is a typical feature of Web 2.0, has been actively studied in the field of learning. Mashups in the field of learning are used to provide tools to support the creation of lecture materials [7] [8], and are used to enhance learners' learning by providing various external content to learners [9].

Hsiao et al. Proposed EduLINK [7] to provide a code similar to the example needed by a teacher when a lecturer (teacher) constructs a lecture on the Java programming language (Fig. 4) based on the Java ontology. By making it possible to reuse the generated web contents, it is possible to reduce the time and labor of the instructor who is required to write lecture data.

Capuano et al. [8] analyzed cognitive processes needed in class, and proposed a template to support the preparation of lecture data suitable for each cognitive process. For example, when the instructor needs to define a new concept for teaching materials, the 'concept definition' template provides a meta search result of the web contents on the Wiktionary site (FIG. 5) And labor. Fig. 6 is a part of lecture data prepared using the template proposed by Capuano et al.

Re-Mashed proposed by Drachsler et al. [9] is a mashup system for PLE (Personal Learning Environment) (Fig. 7). Collaborative filtering of learners' And to recommend meaningful contents to individual learners afterwards.

Although there are various mashup researches in the field of learning, it is limited to the research to reduce the labor force required to write lecture data of the instructor or to provide the external contents to the learner, There is little research on the mashup system to provide the overall knowledge flow or knowledge structure.

[1] C. Seeberg, A. Steinacker, K. Reichenberger, S. Fischer, R. Steinmentz, "Individual Tables of Contents in Web-based Learning Systems" Conference on Hypertext and Hypermedia, pp. 167-168, 1999

[2] M. Thuring, J. Hannemann, J. Haake, "Hypermedia and Cognition: Designing for Comprehension", Designing Hypermedia Application, Communications of the ACM, pp. 57-66, 1995

[3] E. Egan, R. Remde, K. Landauer, C. Lochbaum, M. Gomez, "Behavioral Evaluation and Analysis of a Hypertext Browser", CHI, pp. 205-210, 1989

[4] R. Floyd, M. Jones, D. Rathi, M. Twidale, "Web Mash-ups and Patchwork Prototyping: User-driven Technological Innovation with Web 2.0 and Open Source Software" .1-10, 2007

[5] R. Remde, M. Gomez, K. Landauer, "SuperBook: An Automatic Tool for Information Exploration -Hypertext?" Hypertext, pp. 175-188, 1987

[6] M. D. Merrill, "Knowledge Objects and Mental Models, Advanced Learning Technologies, pp. 244-246, 2000

[7] L. Hsiao, Q. Li, Y. Lin, "Educational Social Linking in Example Authoring", Hypertext, pp. 292-230, 2008

[8] N. Capuano, A. Pierri, F. Colace, M. Gaeta, "A Mash-up Authoring Tool for e-Learning based on Pedagogical Templates", Multimedia Technologies for Distance Learning, pp.1-8, 2009

[9] H. Drachsler, D. Pecceu, T. Arts, E. Hutten, L. Rutledge, P. Rosmalen, H. Hummel, R. Koper, "ReMashed Recommendations for Mash-Up Personal Learning Environments" Synergy of Multiple Disciplines, pp. 788-793, 2009

[10] Kim, Donghee, "Design and Construction of Location Contents System Using Map API Mashup", Master Degree Thesis, 2010

[11] Lee, Hye-kyung, "A Study on Map Education Contents Using Google Map API Mashup", Master's Thesis, Ewha Womans University, 2008

[12] Stephanie Lyn Flosi, "comScore Releases May 2010 U.S. Online Video Rankings". COMSCORE, Written June 24, 2010. Retrieved June 27, 2010. http://www.comscore.com/Press_Events/Press_Releases/2010/6/comScore_Releases_May_2010_U.S._Online_Video_Rankings,

[13] Glenn Chapman, "Youtube Serving up Two Billion Videos Daily". AFP, Written March 16, 2010. Retrieved June 1, 2009.

http://www.google.com/hostednews/afp/article/ALeqM5jK4sI9GfUTCKAkVGhDzpJ1ACZm9Q

[14] Wikipedia: http://www.wikipedia.org

[15] Amazon: http://www.amazon.com

[16] Housing Map: http://www.housingmaps.com

[17] Real Estate Auction: http://www.realestateauction.com

[18] eBay: http://www.ebay.com

[19] SKOS: http://www.w3.org/2004/02/skos/core

[20] DC core: http://purl.org/dc/elements/1.1

[21] FOAF: http: //xmlns.com.foaf/0.1

[22] Search for semantic lectures: http://147.46.70.151:7000/bmi/index.htm

[23] Youtube: http://www.youtube.com

[24] Slideshare: http://slideshare.net

[25] Delicious: http://www.delicious.com

[26] Yahoo Pipe: http://pipes.yahoo.com

The object of the present invention is to solve the problems as described above. In order to systematically provide domain knowledge, it is an object of the present invention to extract a table of contents and an index list from publishing books belonging to a domain, provide domain knowledge in the order of contents, A semantic knowledge structure based information search method and system for providing search results searched by indexes.

In order to achieve the above object, according to the present invention, there is provided a method of searching for domain knowledge, which is connected to a search server through a network, receives a table of contents list and an index list of an issued book, (A) a table of contents list and an index list of the publishing book are input, the table of contents list includes a plurality of table entries arranged in order, and the index list includes a plurality of indexes ; (b) retrieving, from the index list, indexes belonging to the table of contents item; (c) searching the search sites using the searched indexes as keywords; (d) providing a search result of the search site as a search result of the searched index; And (e) providing search results of the indexes belonging to the table of contents item in the order of the table of contents items in the table of contents list.

According to another aspect of the present invention, there is provided a semantic knowledge structure-based information retrieval method using a table of contents and an index, the method comprising: extracting a page range of the table of contents item in the step (b) And searching for an index belonging to the table of contents item.

According to another aspect of the present invention, there is provided a semantic knowledge structure based information retrieval method using a table of contents and an index, wherein, in the step (e), if the word of the index is a word included in the table of contents, The search result of the index is given as a top priority.

As described above, according to the method and system for searching information based on the semantic knowledge structure using the table of contents and the index of the book according to the present invention, It is possible to acquire knowledge more easily.

The results of the experiments on the hypothesis proposed in the present invention are summarized as follows.

1. For beginners of bioinformatics, searching for a disease is effective in understanding knowledge structure related to bioinformatics.

2. For those who are familiar with bioinformatics, searching for the book is effective in understanding the knowledge structure related to bioinformatics.

3. For bioinformatics beginners, searching for a disease is effective in understanding concepts related to bioinformatics.

4. For beginners of bioinformatics, searching for a disease is effective in understanding concepts related to bioinformatics.

5. Satisfaction with the knowledge structure-based retrieval in Hebei showed high satisfaction irrespective of bioinformatics beginner and familiar.

6. Satisfaction with the related concept-based search in the Hebuk area was significantly higher than those who were familiar with bioinformatics.

If the table of contents and index data are gradually opened and shared, such as author, ISBN, which is basically provided with respect to books, the proposed system of the present invention can be guaranteed the diversity and diversity of the field. Also, as the semantic information representation about web contents is gradually spreading, the quality of web contents provided by mashing up in the system can be guaranteed.

FIG. 1 is an interface screen diagram of a SuperBook.
2 shows a housing map.
Fig. 3 is a real estate auction.
4 is an EduLink interface.
Figure 5 shows the learning attributes and related Web 2.0 services.
Fig. 6 is a learning material using a pedagogical template.
7 shows the ReMashed interface.
Figure 8 is a partial table of contents provided by Amazon.
FIG. 9 shows part of the book data entered in TopBraid.
Figure 10 is a partial table of contents converted to RDF.
11 is an attribute created to express the contents information and relationship.
FIG.
Fig.
Fig. Mashup interface for pipes.
15 is a search interface of a semantic lecture.
16 shows a book list providing screen.
17 shows a search screen based on a semantic knowledge structure.
18 shows an index based search screen.
19 is a conceptual network diagram.
20 is a conceptual network-based search screen;

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, the present invention will be described in detail with reference to the drawings.

In the description of the present invention, the same parts are denoted by the same reference numerals, and repetitive description thereof will be omitted.

In order to create a semantic knowledge structure and conceptual network for bioinformatics, we extracted the basic information, table of contents and index information of the bioinformatics book provided in the Amazon [15] book section. The extracted book data (FIG. 8) is semiautomatically generated in the RDF format (FIG. 10) using TopBraid tool (FIG. 9) supporting creation and management of domain ontologies based on RDF (Resource Description Framework) schema and OWL ).

(Simple Knowledge Organization System) [19], DCMI (Dublin Core Metadata Initiative) [20], and FOAF (Friend of A Friend) [20] to represent basic information such as book authors, publishers, ISBNs, 21], and created some attributes for table of contents, index data, and relationship representation. Fig. 11 shows part of the attribute generated in the Hebuk, which is the semantic knowledge structure based information retrieval system proposed in the present study, and Fig. 12 shows the schema of the Hebuk system.

Each description of the specification inherited from the north is as follows.

SKOS: Simple Knowledge Organization System

SKOS is an RDF glossary for representing knowledge vocabulary such as thesaurus, classification system, and so on. Because SKOS was created based on RDF, the knowledge vocabulary system represented by SKOS is machine-readable, interoperable between applications, and can be published to the Web .

l DC: Dublin Core Metadata Initiative (DCMI)

Dublin Core is one of the metadata standards created to support interoperability between metadata, and is metadata created to support meaningful information retrieval and processing. Dublin Core is widely used because it can be easily applied to complex media such as text, images, web pages, sound, and video, and usually uses XML (extensible Markup Language) and RDF.

l FOAF: Friend of A Friend

FOAF is an ontology that can be understood by machines. It is created to describe information about people, relationships between people and people, and relationships between people and objects. FOAF uses RDF and OWL expressions.

12 shows the relationship between a book, a chapter, an index, and a concept. The concept represents a set of keywords having the wb: hasConcept, wb: hasKeyword attribute values extracted from the table of contents and index.

Semantic knowledge structure based on bio-informatics learners' knowledge structure to support the acquisition of related knowledge in a comprehensive and systematic knowledge structure in diverse and abundant web contents sporadically generated by many users as user participation service develops Fig. 13 shows a process of a high-level information search service based on the concept network.

This process can be divided into two phases: knowledge structure generation and mashup. Since the step of generating the knowledge structure with RDF has been described above,

By providing a semantic knowledge structure based on the systematic and general knowledge structure generated by bioinformatics experts, we can mash up a specific web service in semantic knowledge structure, and provide web search results based on semantic knowledge structure to learners.

The mashup with the North has been limited to web services such as Wikipedia [14], YouTube [23], SlideShare [24], and Delicious [25], in which web contents that are easy to use as learning materials and lecture materials are generated. The mashup was implemented using Yahoo! Pipes [26], one of the mashup support services.

Yahoo Pipe (Figure 14) is more developer-oriented compared to other mashup support services such as Google's Mashup Editor or Microsoft's Popfly

It is data-oriented, which is inconvenient for general users to write, but has the advantage of achieving much more flexible results.

Curriculum Development for Bioinformatics based on the KRF-2008-D00035 Research Support Semantic Lecture Search [22] By providing services and mash-ups, students can learn bioinformatics related lectures, Lecture materials, lecture plans, etc. (Fig. 15).

It consists of a semantic knowledge structure based search interface with the main purpose of learning support for bioinformatics beginners and a conceptual network based search interface with the main purpose of learning support of bioinformatics familiar.

In order for a learner to use the bookbook, he first selects an interest book from the list of bioinformatics books provided in the bookbook (Fig. 16), and then, based on the learner's knowledge on bioinformatics, .

- Semantic knowledge structure based search interface

Semantic knowledge structure based retrieval is mainly aimed at beginners of bioinformatics (bioinformatics beginner) who do not have background knowledge about bioinformatics, and knowledge structure of one book is provided.

When a learner looks at the knowledge structure and selects a term that he or she is interested in, he or she will receive search results of relevant web contents from Wikipedia, Delicious, SlideShare, and YouTube through a mashup. At the same time, (Fig. 17).

In addition to providing indexes related to the table of contents, a richer knowledge structure can be provided (FIG. 17), and mashup results through indexes are also provided (FIG. 18).

- Conceptual network-based search interface

The concept of network-based retrieval is mainly aimed at learner (bioinformatics familiar) who has a background knowledge about bioinformatics and provides concepts related to one bioinformatics concept together in network form.

When a learner selects a term of interest by looking at a knowledge structure or an index, groups of concepts related to the concept are grouped and provided (FIG. 19) so that the learner can easily grasp the related concept (FIG.

In addition, by adding the conceptual search function, the learner supports not only the knowledge structure or index based search but also free search.

Although the present invention has been described in detail with reference to the above embodiments, it is needless to say that the present invention is not limited to the above-described embodiments, and various modifications may be made without departing from the spirit of the present invention.

In order to systematically provide domain knowledge, the present invention provides a semantic knowledge structure structure that extracts a table of contents and a list of indexes from publishing books belonging to a domain, provides domain knowledge in the order of a table of contents and provides a search result with an index belonging to a table of contents It is useful for developing information retrieval systems.

Claims (3)

A semantic knowledge structure based information retrieval method using a table of contents and an index of a book which is connected to a search server through a network and receives a list of contents and an index list of an issuance book and retrieves domain knowledge in the order of the list of contents,
(a) receiving a table of contents list and an index list of the publishing book, the table of contents list including a plurality of table entries arranged in order, the index list including a plurality of indexes;
(b) retrieving, from the index list, indexes belonging to the table of contents item;
(c) searching the search server using the searched indexes as a keyword;
(d) providing a search result of the search server as a search result of the searched index; And
(e) providing search results of indexes belonging to the table of contents items in the order of the table of contents items in the table of contents list,
The method according to claim 1, wherein, in the step (b), a page range of the table of contents item is extracted, and the indexes having pages belonging to the page range are retrieved with an index belonging to the table of contents item. Based information retrieval method.
delete The method according to claim 1,
Wherein if the word of the index is a word included in the table of contents, the search result of the index is provided as a top priority among the search results of the table of contents item. Semantic knowledge structure based information retrieval method.

Images