![ISSN: 2278 – 1323
International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
Volume 2, Issue 2, February 2013
677
All Rights Reserved © 2013 IJARCET
Schema can be viewed as a essential language for writing
ontologies. But there is a need for more powerful ontology
languages which can expand RDF Schema and allow the
representations of more complex relationships between Web
objects.
OWL (Web Ontology Language) is an ontology language
that extends expressiveness of RDFS. OWL is a declarative
knowledge representation language which formally defines
meaning for creating ontology. OWL has three sublanguages
like OWL-Lite, OWL-DL, and OWL-Full. OWL-Lite and
OWL-DL are based on Description logics.
II. INTRODUCTION TO INFERENCE ENGINE
Inference on semantic web is used derive a new
relationship. For the semantic web, inference is a process to
infer a new relationship from existing resources and some
addition information in form of “set of rules”. From some
addition information can be define to given sources using
vocabularies or rules set. Inference is requiring for
processing given knowledge available on semantic web.
Inference base technique is also use to check data
inconsistency at time of data integration.
The inference engine can be described as a form of finite
state machine with a cycle consisting of three action states:
match rules, select rules, and execute rules [1]. In the first
match rules state, the inference engine finds all of the rules
that are predefined in database. Then in next step the
inference engine passes the data set to the second state which
known as select rules. Finally the selected databases are
passed to the third state which is known as execute rules. The
inference engine executes the selected rules, with the
selected data items as parameters.
Example:
Here, we have following fact:
A is parent of C,
B is parent of C,
A is men,
B is Women
Based on given fact, we derive new information is that “B is
wife of A”.
Inference engine derive a new fact based on existing fact
by using some sets of rules. Rules are very important for
controlling the steps of inference process. The inference rules
are specifying by means of ontology language and
description language. Reasoners use some logic to perform
reasoning task like Description logic, First Order logic
(FOL), Predicate logic and Propositional logic.
Inference engine can be used for reasoning task for
Ontologies and their instances. Based on, it derives a new
knowledge based on existing knowledge available on web.
Most of the inference engine is use query language for
reasoning task. In Current world, it widely use in knowledge
engineering and artificial intelligence.
There are three alternative strategies to manage inferred
knowledge:
Forward Chaining Method:
In forward chaining method it starts with the
available data and uses inference rules to extract
more data until a goal is reached.
Backward Chaining Method:
Backward chaining starts with a list of goals and
works backwards from the data available.
Hybrid Method:
This method is more complex so it is widely use in
expert system. In this method, the rules of both
forward and backward chaining can apply.
III. INFERENCE ENGINES
INFERENCE ENGINES In our comparative analysis we
have studied following
1)PELLET
Pellet is an OWL DL reasoner for semantic web which
uses tableaux algorithms for development of DL description
logics and decision procedure. It supports the full
expressivity OWL DL including reasoning, SROIQ and also
SWRL rules. Library of Jena and OWL API is used for
conjunction. Pellet API provides functionalities to see the
species validation, check consistency of ontologies, classify
the taxonomy, check entailments and answer a subset of
RDQL queries.
2) FACT
FaCT is known as "Fast Classification of Terminologies" .
It is a Description Logic (DL) classifier that is used for modal
logic satisfiability testing. The most interesting features of
the FaCT are its expressive logic, SHIQ is sufficiently
expressive to be used as a reasoner for the Dedicated Logic
Register (DLR) logic; and its support for reasoning with
arbitrary knowledge bases and other is its optimised tableaux
implementation and its Common Object Request Broker
Architecture - CORBA based client-server architecture.
3) FACT++
FaCT++ an improved version of FaCT employs
tableaux algorithms for SHOIQ(D) description logic that
support general A-boxes and T-boxes and implemented in
C++ but has very limited user interface and services as
compared to other reasoner[2]. It supports the OWL-API, the
lisp-API and the DIG interface.
4) RACERPRO
Racerpro is lips based OWL DL reasoner. It supports
expressivity of SHIQ. It implements a tableau-based decision
procedure for general T-Boxes (subsumption, satisfiability,
classification) and A-Boxes (retrieval, nRQL query
answering). It supports the OWL-API and the DIG interface.
It follows the multiple optimization strategies for better](https://image.slidesharecdn.com/ijarcet-vol-2-issue-2-676-678-130822074939-phpapp02/85/Ijarcet-vol-2-issue-2-676-678-2-320.jpg)
![ISSN: 2278 – 1323
International Journal of Advanced Research in Computer Engineering & Technology (IJARCET)
Volume 2, Issue 2, February 2013
All Rights Reserved © 2013 IJARCET
678
reasoning, etc. support including dependency-directed
backtracking, and transformation of axioms, model caching
and merging.
5) KAON2
KAON2 is free java based reasoner. It is based on
OWL-DL and Frame Logic. It is an infrastructure of
managing OWLDL, SWRL and F-Logic ontologies. It
supports answering conjunctive queries, although without
true non-distinguished variables. KAON2 is a successor of
KAON project used extension of RDFS [26]. It implements a
resolution-based decision procedure for general TBoxes
(subsumption, satisfiability, classification) and ABoxes
(retrieval, conjunctive query answering). It support Java
based interface and DIG interface.
6) HERMIT
HermiT is a free java based OWL reasoner that use novel
“hyper-tableau" calculus [3]. Hermit reasoner employs
reasoning on SHIQ (D). It is available free for
non-commercial usage. Takes OWL file as input and perform
various reasoning tasks like consistency checking, identify
subsumption relationships between classes and more. It also
computes partial order of classes occurring in OWL. It is
different from other reasoner like Pellet and Fact such a way
that it implements hyper-tableau for SHIQ(D) which is much
less non-deterministic than the existing tableau algorithms. It
supports the interface of OWL API.
IV. CONCLUSION
Based on findings to understand and use the semantic data on
the web there is a requirement of inference engine. This
paper describes comparison of various inference engines on
different measurement criteria. Here we have tried to be
exhaustive in comparison but there may be other engines
which we have not discussed. Our comparative analysis table
may be further helpful in selecting inference engines for
semantic web applications and for future research works.
V. ACKNOWLEDGMENT
The success of this research depends largely on the
encouragement and guidelines of many others. we take this
opportunity to express my gratitude to the people who have
been instrumental in the successful completion of this
research. We would like to show my greatest appreciation
to Ass. Prof Priyanka Trikha. Without her encouragement
and guidance this work would not have materialized.
The guidance and support received from all of my friends
and all our faculty members who give their valuable
suggestions to me for this research. I am grateful for their
constant support and help.
REFERENCES
[1] A Comparative Study Of Inference Engines by
Swapna Singh, Ragini Karwayun, 7th International
conference on Information Technology, 2010.
[2] http://kaon2.semanticweb.org
[3] http://owl.man.ac.uk/hoolet/
[4]http://www2012.wwwconference.org/about/submit/
call-for-papers/semantic-web/
[5] “Reducing OWL entailment to description logic
satisfiability” Ian Horrocks and Peter Patel-Schneider
Journal of Web Semantics, 1(4):345{357, 2004.}
[6] Tim Berners-Lee, James Hendler, and Ora Lassila.
The semantic web. Scienti_c American, 284(5):34{43,
2001.
[7] B. Parsia, E. Sirin, Pellet: An owl dl reasoner. In: Proc.
International Semantic Web Conference. (2005).
[8] Racer: A Core Inference Engine for the Semantic Web
by Volker Haarslev† and Ralf M¨oller., 2003.
[9] F-OWL: an Inference Engine for the Semant ic Web by
Youyong Zou, Tim Finin and Harry Chen, 2003
[10] Bossam: An Extended Rule Engine for OWL
Inferencing by Minsu Jang and Joo-Chan Sohn, 2004
[11] FIRE – A Description logic Based Rule Engine for
OWL Ontologies with SWRL-like Rules by Kruthi
Bhupalam, 2005
[12] OWLer : A semantic web ontology inference
engine,by young-ku-ha, in 7th International
conference on advanced communication
tecchnology,2005
[13] Maryland Information and Network Dynamics Lab.:
Pellet OWL Reasoner. (2003)](https://image.slidesharecdn.com/ijarcet-vol-2-issue-2-676-678-130822074939-phpapp02/85/Ijarcet-vol-2-issue-2-676-678-3-320.jpg)
Ijarcet vol-2-issue-2-676-678
- 1. ISSN: 2278 – 1323 International Journal of Advanced Research in Computer Engineering & Technology (IJARCET) Volume 2, Issue 2, February 2013 All Rights Reserved © 2013 IJARCET 676 Abstract— Semantic web is a web of data, where data should be related to one another and also Knowledge will be organized in conceptual spaces according to its meaning. To understand and use the data and knowledge encoded in semantic web documents requires inference engine. There are number of inference engines used for consistency checking and classification like Pellet, Fact, Fact++, Hermit, Racer Pro, KaON2, and Base Visor. Some of them are reviewed and tested for few prebuilt ontologies. This paper presents the analysis of different inference engines with set of ontologies. It requires assessment and evaluation before selecting an appropriate inference engine for a given application. Keywords—Description logic, HTML, Inference, OWL, RDF, RDFS. I. INTRODUCTION Today’s Web uses the hyperlinks that define relationship between current page and target page whereas semantic web define relationships among data on web has motivate creation of new technologies and standards to analyze and understand large amount of data on web and infer new knowledge. Semantic Web is an emerging technology regarded as the next generate on Web paradigm providing machine understandable information that is based on meaning (Tim Berners-Lee 2001). It has been recognized as a promising new area of research to define and describe relationships among data (resources) on the Web. Semantic Web is an extension of Web 2.0. Today’s Web which is used for coding a body of text with images is mainly written in Hypertext Markup Language (HTML). Whereas The Semantic Web uses languages such as Resource Description Framework (RDF), Web Ontology Language (OWL) and Extensible Markup Language (XML) specifically designed for data that can describe arbitrary things. The Web Ontology Language (OWL) is an approach for knowledge representation which describes basic concepts and defines relationships among data. The Semantic Web is generally built on syntaxes which use URIs to represent data, usually in triples based structures; these syntaxes are called RDF (Resource Descriptive Framework). Figure 1: Semantic web architecture The semantic web architecture defined by W3C is of step by step layered model. RDF, RDFS, OWL are standards of semantic web. RDF Resource Description Framework is a basic data model, like the entity- relationship model, for writing simple statements about Web objects (resources). It is a semantic web standard for defining information on the Web. RDF provides the language for expressing the meaning of terms and concepts in a form that machines can readily process. RDF statement contains 3 main parts: Subject, Property, and Object, respectively. Subject that identifies things for that statement, Predicate that identifies property/characteristics of the subject specifies and Object that identifies value of that property. In terms of the simple graph above, the: Subject is the T-shirt Predicate (property) is the color Object is yellow RDF does not make assumptions about any particular domain and does not define semantics of any domain. This can be done using RDFS. RDFS (Resource Descriptive Framework Schema) is an extension of RDF in which it allows to describing taxonomies of classes and properties. It also allows the description of class hierarchy and property hierarchy. RDF Interpreting Inference Engine for Semantic Web MS. PRIYANKA PATEL #1 , MS. PRIYANKA TRIKHA #2 COMPUTER SCIENCE AND ENGINEERING DEPARTMENT SRI BALAJI COLLEGE OF ENGINEERING & TECHNOLOGY Jaipur-302012 T-Shirt Yellow COLOR
- 2. ISSN: 2278 – 1323 International Journal of Advanced Research in Computer Engineering & Technology (IJARCET) Volume 2, Issue 2, February 2013 677 All Rights Reserved © 2013 IJARCET Schema can be viewed as a essential language for writing ontologies. But there is a need for more powerful ontology languages which can expand RDF Schema and allow the representations of more complex relationships between Web objects. OWL (Web Ontology Language) is an ontology language that extends expressiveness of RDFS. OWL is a declarative knowledge representation language which formally defines meaning for creating ontology. OWL has three sublanguages like OWL-Lite, OWL-DL, and OWL-Full. OWL-Lite and OWL-DL are based on Description logics. II. INTRODUCTION TO INFERENCE ENGINE Inference on semantic web is used derive a new relationship. For the semantic web, inference is a process to infer a new relationship from existing resources and some addition information in form of “set of rules”. From some addition information can be define to given sources using vocabularies or rules set. Inference is requiring for processing given knowledge available on semantic web. Inference base technique is also use to check data inconsistency at time of data integration. The inference engine can be described as a form of finite state machine with a cycle consisting of three action states: match rules, select rules, and execute rules [1]. In the first match rules state, the inference engine finds all of the rules that are predefined in database. Then in next step the inference engine passes the data set to the second state which known as select rules. Finally the selected databases are passed to the third state which is known as execute rules. The inference engine executes the selected rules, with the selected data items as parameters. Example: Here, we have following fact: A is parent of C, B is parent of C, A is men, B is Women Based on given fact, we derive new information is that “B is wife of A”. Inference engine derive a new fact based on existing fact by using some sets of rules. Rules are very important for controlling the steps of inference process. The inference rules are specifying by means of ontology language and description language. Reasoners use some logic to perform reasoning task like Description logic, First Order logic (FOL), Predicate logic and Propositional logic. Inference engine can be used for reasoning task for Ontologies and their instances. Based on, it derives a new knowledge based on existing knowledge available on web. Most of the inference engine is use query language for reasoning task. In Current world, it widely use in knowledge engineering and artificial intelligence. There are three alternative strategies to manage inferred knowledge: Forward Chaining Method: In forward chaining method it starts with the available data and uses inference rules to extract more data until a goal is reached. Backward Chaining Method: Backward chaining starts with a list of goals and works backwards from the data available. Hybrid Method: This method is more complex so it is widely use in expert system. In this method, the rules of both forward and backward chaining can apply. III. INFERENCE ENGINES INFERENCE ENGINES In our comparative analysis we have studied following 1)PELLET Pellet is an OWL DL reasoner for semantic web which uses tableaux algorithms for development of DL description logics and decision procedure. It supports the full expressivity OWL DL including reasoning, SROIQ and also SWRL rules. Library of Jena and OWL API is used for conjunction. Pellet API provides functionalities to see the species validation, check consistency of ontologies, classify the taxonomy, check entailments and answer a subset of RDQL queries. 2) FACT FaCT is known as "Fast Classification of Terminologies" . It is a Description Logic (DL) classifier that is used for modal logic satisfiability testing. The most interesting features of the FaCT are its expressive logic, SHIQ is sufficiently expressive to be used as a reasoner for the Dedicated Logic Register (DLR) logic; and its support for reasoning with arbitrary knowledge bases and other is its optimised tableaux implementation and its Common Object Request Broker Architecture - CORBA based client-server architecture. 3) FACT++ FaCT++ an improved version of FaCT employs tableaux algorithms for SHOIQ(D) description logic that support general A-boxes and T-boxes and implemented in C++ but has very limited user interface and services as compared to other reasoner[2]. It supports the OWL-API, the lisp-API and the DIG interface. 4) RACERPRO Racerpro is lips based OWL DL reasoner. It supports expressivity of SHIQ. It implements a tableau-based decision procedure for general T-Boxes (subsumption, satisfiability, classification) and A-Boxes (retrieval, nRQL query answering). It supports the OWL-API and the DIG interface. It follows the multiple optimization strategies for better
- 3. ISSN: 2278 – 1323 International Journal of Advanced Research in Computer Engineering & Technology (IJARCET) Volume 2, Issue 2, February 2013 All Rights Reserved © 2013 IJARCET 678 reasoning, etc. support including dependency-directed backtracking, and transformation of axioms, model caching and merging. 5) KAON2 KAON2 is free java based reasoner. It is based on OWL-DL and Frame Logic. It is an infrastructure of managing OWLDL, SWRL and F-Logic ontologies. It supports answering conjunctive queries, although without true non-distinguished variables. KAON2 is a successor of KAON project used extension of RDFS [26]. It implements a resolution-based decision procedure for general TBoxes (subsumption, satisfiability, classification) and ABoxes (retrieval, conjunctive query answering). It support Java based interface and DIG interface. 6) HERMIT HermiT is a free java based OWL reasoner that use novel “hyper-tableau" calculus [3]. Hermit reasoner employs reasoning on SHIQ (D). It is available free for non-commercial usage. Takes OWL file as input and perform various reasoning tasks like consistency checking, identify subsumption relationships between classes and more. It also computes partial order of classes occurring in OWL. It is different from other reasoner like Pellet and Fact such a way that it implements hyper-tableau for SHIQ(D) which is much less non-deterministic than the existing tableau algorithms. It supports the interface of OWL API. IV. CONCLUSION Based on findings to understand and use the semantic data on the web there is a requirement of inference engine. This paper describes comparison of various inference engines on different measurement criteria. Here we have tried to be exhaustive in comparison but there may be other engines which we have not discussed. Our comparative analysis table may be further helpful in selecting inference engines for semantic web applications and for future research works. V. ACKNOWLEDGMENT The success of this research depends largely on the encouragement and guidelines of many others. we take this opportunity to express my gratitude to the people who have been instrumental in the successful completion of this research. We would like to show my greatest appreciation to Ass. Prof Priyanka Trikha. Without her encouragement and guidance this work would not have materialized. The guidance and support received from all of my friends and all our faculty members who give their valuable suggestions to me for this research. I am grateful for their constant support and help. REFERENCES [1] A Comparative Study Of Inference Engines by Swapna Singh, Ragini Karwayun, 7th International conference on Information Technology, 2010. [2] http://kaon2.semanticweb.org [3] http://owl.man.ac.uk/hoolet/ [4]http://www2012.wwwconference.org/about/submit/ call-for-papers/semantic-web/ [5] “Reducing OWL entailment to description logic satisfiability” Ian Horrocks and Peter Patel-Schneider Journal of Web Semantics, 1(4):345{357, 2004.} [6] Tim Berners-Lee, James Hendler, and Ora Lassila. The semantic web. Scienti_c American, 284(5):34{43, 2001. [7] B. Parsia, E. Sirin, Pellet: An owl dl reasoner. In: Proc. International Semantic Web Conference. (2005). [8] Racer: A Core Inference Engine for the Semantic Web by Volker Haarslev† and Ralf M¨oller., 2003. [9] F-OWL: an Inference Engine for the Semant ic Web by Youyong Zou, Tim Finin and Harry Chen, 2003 [10] Bossam: An Extended Rule Engine for OWL Inferencing by Minsu Jang and Joo-Chan Sohn, 2004 [11] FIRE – A Description logic Based Rule Engine for OWL Ontologies with SWRL-like Rules by Kruthi Bhupalam, 2005 [12] OWLer : A semantic web ontology inference engine,by young-ku-ha, in 7th International conference on advanced communication tecchnology,2005 [13] Maryland Information and Network Dynamics Lab.: Pellet OWL Reasoner. (2003)