Partha Pakray, International Journal of Computer Science and Mobile Computing, Vol.3 Issue.6, June- 2014, pg. 528-534 © 2014, IJCSMC All Rights Reserved 528 Available Online at www.ijcsmc.com International Journal of Computer Science and Mobile Computing A Monthly Journal of Computer Science and Information Technology ISSN 2320–088X IJCSMC, Vol. 3, Issue. 6, June 2014, pg.528 – 534 RESEARCH ARTICLE Text Grouping using Textual Entailment Partha Pakray Department of Computer & Information Science (IDI) Norwegian University of Science and Technology (NTNU) Trondheim, Norway www.parthapakray.com partha.pakray@idi.ntnu.no Abstract — Textual Entailment is an important field in Natural Language Processing domain. Given two texts called T (Text) and H (Hypothesis), the textual entailment recognition is the task of deciding whether the meaning of H can be logically inferred from that of T. A Textual Entailment (TE) system has developed and this system has tested on various entailment standard datasets. This TE will apply to different texts then the TE system will group them into different single group. A corpus has created for this experiment that has total 10 groups which contains 3540 sentences. F-score of the textual entailment system is 61% and will detect 8 groups correctly from 10 groups. Keywords— Natural Language Processing, Textual Entailment, reverb, Support Vector Machine I. INTRODUCTION Many efforts have devoted by the Natural Language Processing (NLP) community to develop advanced methodologies in Textual Entailment (TE), which is considered as a core NLP task. Various international conferences and several evaluation track competitions on Textual Entailment have been held, notably at PASCAL-Pattern Analysis, Statistical Modelling and Computational Learning 1 , Text Analysis Conferences (TAC) 2 organized by the United States National Institute of Standards and Technology (NIST), Evaluation Exercises on Semantic Evaluation (SemEval) 3 , National Institute of Informatics Test Collection for Information Retrieval System (NTCIR) 4 since 2005. Textual entailment can be more formally defined [1] as  A text T entails a hypothesis H, if H is true in every circumstance in which T is true.  A text T entails a hypothesis H if, typically, a human reading T would infer that H is most likely true. For example, the text T = “John’s assassin is in jail” entails the hypothesis H = “John is dead”; indeed, if there exists one’s assassin, then this person is dead. Similarly, T = “Mary lives in France” entails H = “Mary lives in Europe”. On the other hand, T = “Mary lives in Europe” does not entail H = “Mary lives in US”. Main focus of this experiment is that Text Grouping (i.e. clustering) can do by Textual Entailment. For this experiment own developed TE system used that already developed previously and participated various Recognising Textual Entailment (RTE) Challenges and tested on RTE datasets. This TE system has successfully applied to Question Answering (QA) domain and participated QA track (QA4MRE) at Conference and Labs of 1 http://pascallin.ecs.soton.ac.uk/Challenges/ 2 http://www.nist.gov/tac/tracks/index.html 3 http://semeval2.fbk.eu/semeval2.php 4 http://research.nii.ac.jp/ntcir/ntcir-9/