Welcome to Scribd, the world's digital library. Read, publish, and share books and documents. See more
Download
Standard view
Full view
of .
Save to My Library
Look up keyword
Like this
1Activity
0 of .
Results for:
No results containing your search query
P. 1
A Domain Specific Ontology Based Semantic Web Search Engine

A Domain Specific Ontology Based Semantic Web Search Engine

Ratings: (0)|Views: 75 |Likes:
Published by Phuong Nguyen

More info:

Published by: Phuong Nguyen on Jun 05, 2011
Copyright:Attribution Non-commercial

Availability:

Read on Scribd mobile: iPhone, iPad and Android.
download as PDF, TXT or read online from Scribd
See more
See less

06/05/2011

pdf

text

original

 
 
A Domain Specific Ontology BasedSemantic Web Search Engine
Debajyoti Mukhopadhyay
1, 2
Aritra Banik
1
Sreemoyee Mukherjee
1
Jhilik Bhattacharya
1
 1
Web Intelligence & Distributed Computing Research Lab,
Techno India,West Bengal University of Technology
 
EM 4/1, Salt Lake Sector V, Calcutta 700091, India
 
debajyoti.mukhopadhyay@gmail.com
Young-Chon Kim
2
2
Advanced Communications & Networks Lab, Division of Electronics & Information EngineeringChonbuk National University561-756 Jeonju, Republic of Korea
 ABSTRACT 
 
Since its emergence in the 1990s the World WideWeb (WWW) has rapidly evolved into a hugemine of global information and it is growing insize everyday. The presence of huge amount of resources on the Web thus poses a serious problem of accurate search. This is mainlybecause today’s Web is a human-readable Webwhere information cannot be easily processed bymachine. Highly sophisticated, efficient keyword based search engines that have evolved todayhave not been able to bridge this gap. So comesup the concept of the Semantic Web which isenvisioned by Tim Berners-Lee as the Web of machine interpretable information to make amachine processable form for expressinginformation. Based on the semantic Webtechnologies we present in this paper the designmethodology and development of a semanticWeb search engine which provides exact searchresults for a domain specific search. This searchengine is developed for an agricultural Websitewhich hosts agricultural information about thestate of West Bengal.
Keywords
 
Search engine, Keyword based search, SemanticWeb, Ontology, Resource DescriptionFramework (RDF)
 
1. INTRODUCTION
A
search engine
is a document retrieval systemdesigned to help find information stored in acomputer system, such as on the World WideWeb, inside a corporate or proprietary network,or in a personal computer. The search engineallows one to ask for content meeting specificcriteria (typically those containing a given wordor phrase) and retrieves a list of items that matchthose criteria. Regardless of the underlyingarchitecture, users specify keywords that matchwords in huge search engine databases,producing a ranked list of URLs and snippets of Web-pages in which the keywords matched.Although such technologies are mostly used,users are still often faced with the daunting task of sifting through multiple pages of results, manyof which are irrelevant.Surveys indicate that almost 25% of Web searchers are unable to find useful results inthe first set of URLs that are returned [6].Tim Berners-Lee, the inventor of theWorld Wide Web, defines the
Semantic Web
as“The Web of data with meaning in the sense thata computer program can learn enough aboutwhat the data means [in order] to process it” [1].Rather than a Web filled only withhuman-interpretable information, Berners-Lee’svision includes an extended Web thatincorporates
machine interpretable
information,enabling machines to process the volumes of 
 
 available information, acting on behalf of theirhuman counterparts [7].In this paper, we discuss the basic ideaof the semantic Web and describe a design anddevelopment methodology for a domain specificsemantic Web search engine based on ontologymatching w
hich
not only overcomes theproblem of knowledge overhead but alsosupports complex queries. Further, it is able toproduce exact answers that in one hand satisfyuser queries and on the other hand are self-explanatory and understandable by end users.
2. SEMANTIC WEB SEARCH ENGINE2.1 The working of a regular search engine
For most internet users, a search engine is thestarting point of finding desired information inthe Web. The most common form of text searchused by the majority of popular search engineson the Web is
keyword based search
that is
,
 they do their text query and retrieval usingkeywords. The working of any regular searchengine may be summarized as follows:
 
 
Search engine searches its enormousdatabase for the keyword - entered bythe user (after pressing the searchbutton.)
 
Every engine has its own collectionsystem to fill its database.
 
Indexing system is used to organize thedatabase - permits faster searching
 
Returns a list of hit -includes relevant(as well as irrelevant) pagesThis keyword based search technique givesrise to several problems listed as follows:
 
The Web is growing much faster thanany present-technology search enginecan possibly index. In 2006, some usersfound major search-engines becameslower to index new Web-pages.
 
Keyword searches have a tough timedistinguishing between words that arespelled the same way, but meansomething
 
different. This often resultsin hits that are
 
completely irrelevant tothe query.
 
Some search engines also have troublewith stemming, i.e., if the word "big,"is entered, should it return a hit on theword, "bigger?" What about singularand plural words? What about verbtenses that differ from the wordsomeone entered by only an "s," or an"ed"?
 
Search engines also cannot return hitson keywords that mean the same, butare not actually entered in the query. Aquery on heart disease would not returna document that used the word "cardiac"instead of "heart."
 
Users are returned thousands to millionsof Web pages in return of their queries,of which majority prove to be irrelevantto the query submitted and is impossiblefor any user to go through.In view of the above mentioned problems,come up the concept of semantic Web andsemantic Web search engines.
2.2
 
Semantic Web and Semantic SearchEngine
“The Semantic Web is the representation of dataon the World Wide Web. It is a collaborativeeffort led by W3C with participation from a largenumber of researchers and industrial partners. Itis based on the Resource Description Framework (RDF), which integrates a variety of applicationsusing XML for syntax and URIs for naming.” –W3C Semantic Web. The Semantic Web is aframework that allows publishing, sharing, andreusing data and knowledge on the Web andacross applications, enterprises, and communityboundaries [4].
 
Currently, the Semantic Web,consisting of Semantic Web documents typicallyencoded in the languages RDF and OWL, isessentially a Web universe parallel to the Web of HTML documents [5]. Knowledge encoded inSemantic Web languages such as RDF differsfrom both the largely unstructured free textfound on most Web pages and the highlystructured information found in databases. Suchsemi-structured information requires using acombination of techniques for effective indexingand retrieval. RDF and the Web OntologyLanguage (OWL) which are ontology basedprocedures or representing knowledge on theWeb, introduce aspects beyond those used inordinary XML, allowing users to define terms
 
 (for example, classes and properties), expressrelationships among them, and assert constraintsand axioms that hold for well-formed data. Anapplication of the emerging Semantic Web is aSemantic Web search engine which searches theSemantic Web documents against a user queryfor accurate results. Our work uses RDF encodedSemantic Web documents which are searched inresponse to a user query for exact results.
3. ONTOLOGY
An ontology is an explicit specification of sometopic. For our purposes, it is a formal anddeclarative representation which includes thevocabulary (or names) for referring to the termsin that subject area and the logical statementsthat describe what the terms are, how they arerelated to each other, and how they can or cannotbe related to each other. Therefore, Ontologyprovides a vocabulary for representing andcommunicating knowledge about some topic anda set of relationships that hold among the termsin that vocabulary [2] [3].
Why develop an Ontology?
 
To enable a machine to use theknowledge in some application.
 
To enable multiple machines to sharetheir knowledge.
 
To help yourself understand some areaof knowledge better.
 
To help other people understand somearea of knowledge.
 
To help people reach a consensus intheir understanding of some area of knowledge.In our project we used Resource DescriptionFramework or RDF to represent knowledge. Forexample, if we need to describe a subject interms of its classes and their relationships usingRDF, we are creating an Ontology.As our project deals with the crops domain,the designed ontology is shown in Figure 1. InFigure 2, the general information ontology isdepicted. A relational diagram is shown inFigure 3 to depict some classes, instances, andrelations among them in the crops domain.CropsVegetables Fruits Cereals
potato apple ricepumpkin orange wheat
instances
.... ..... .....
Figure 1.
The Crops ontology: Crops is the superclassof its subclasses vegetables, fruits, and cereals
Generalinformation
Season FertilizerCost
MarketLocation
Figure2.
The general information ontology: Generalinformation is the superclass of subclasses: season,fertlizer, cost, market location
iof 
 
 
CropsGeneralInformationMango
K123
Fertilizer reqd formangoSubclass of 
Fruit
Subclass of 
Fertilizer
iof iof 
Figure 3.
Some classes, instances, and relationsamong them in the crops domain; shaded boxes areused to indicate
classes
and non-shaded for
instances
;iof indicates
instance of 
 

You're Reading a Free Preview

Download
/*********** DO NOT ALTER ANYTHING BELOW THIS LINE ! ************/ var s_code=s.t();if(s_code)document.write(s_code)//-->