Eyal Shnarch   -   אייל שנרך
Department of Computer Science, Bar Ilan University

[Contact | Publications | Downloads | Teaching | Talks ]

Eyal Shnarch Contact Information

Department of Computer Science
Bar Ilan University,
Ramat Gan, 52900, Israel
Phone: +972-3-5318054
Email: shey @ cs.biu.ac.il
How to pronounce my name

I am a research staff member at the Machine Learning for Text Analytics group of IBM Haifa Research Lab. My research field is Natural Language Processing (NLP). Here is my homepage at IBM.

Prior to IBM research, I completed my PhD at the Department of Computer Science, Bar-Ilan University, under the supervision of Ido Dagan and Jacob Goldberger.
I was a part of the NLP lab at Bar Ilan university. My main research interest was semantic Inference. I developed algorithms for the acquisition of lexical semantic relations, proposed methods for the integration of knowledge resources and built probabilistic models for inference at the lexical level.


  1. Eyal Shnarch, Erel Segal Halevi, Jacob Goldberger, Ido Dagan. PLIS: a Probabilistic Lexical Inference System. ACL (Demonstration), 2013. PDF poster code demo
  2. Eyal Shnarch, Ido Dagan, Jacob Goldberger. A Probabilistic Lexical Model for Ranking Textual Inferences. *SEM (Joint Conference on Lexical and Computational Semantics), 2012. PDF slides code
  3. Asher Stern, Amnon Lotan, Shachar Mirkin, Eyal Shnarch, Lili Kotlerman, Jonathan Berant, and Ido Dagan. Knowledge and Tree-Edits in Learnable Entailment Proofs. TAC, 2011. PDF
  4. Eyal Shnarch, Jacob Goldberger, Ido Dagan. Towards a Probabilistic Model for Lexical Entailment. TextInfer, 2011. PDF slides
  5. Eyal Shnarch, Jacob Goldberger, Ido Dagan. A Probabilistic Modeling Framework for Lexical Entailment. ACL, 2011. PDF
  6. Asher Stern, Eyal Shnarch, Amnon Lotan, Shachar Mirkin, Lili Kotlerman, Naomi Zeichner, Jonathan Berant, and Ido Dagan. Rule Chaining and Approximate Match in textual inference. TAC, 2010. PDF
  7. Shachar Mirkin, Jonathan Berant, Ido Dagan and Eyal Shnarch. Recognising Entailment within Discourse. COLING, 2010. PDF
  8. Shachar Mirkin, Roy Bar-Haim, Jonathan Berant, Ido Dagan, Eyal Shnarch, Asher Stern and Idan Szpektor. Addressing Discourse and Document Structure in the RTE Search Task. TAC, 2009. PDF
  9. Eyal Shnarch, Libby Barak, Ido Dagan. Extracting Lexical Reference Rules from Wikipedia. ACL, 2009. PDF slides db (for a new version see WikiRules! 2.0 below)
  10. Libby Barak, Ido Dagan, Eyal Shnarch. Text Categorization from Category Name via Lexical Reference. NAACL, 2009. PDF
  11. Shachar Mirkin, Ido Dagan, Eyal Shnarch. Evaluating the Inferential Utility of Lexical-Semantic Resources. EACL, 2009. PDF
  12. Roy Bar-Haim, Jonathan Berant, Ido Dagan, Iddo Greental, Shachar Mirkin, Eyal Shnarch and Idan Szpektor. Efficient Semantic Deduction and Approximate Matching over Compact Parse Forests. TAC, 2008. PDF
  13. Idan Szpektor, Eyal Shnarch and Ido Dagan. Instance-based Evaluation of Entailment Rule Acquisition. ACL, 2007. PDF
  14. Roy Bar-Haim, Ido Dagan, Iddo Greental and Eyal Shnarch. Semantic Inference at the Lexical-Syntactic Level. AAAI, 2007. PDF
  15. Oren Glickman, Ido Dagan and Eyal Shnarch. Lexical Reference: a Semantic Matching Subtask. EMNLP, 2006. PDF


  • PLIS: a Probabilistic Lexical Inference System
    An open source Probabilistic Lexical Inference System which acquires and integrates lexical knowledge from various input resources and provides several semantic inference models to estimate entailment probability.
    Developers of text processing applications, use PLIS to:
    • overcome lexical variability
    • incorporate background knowledge
    • as a lightweight inference system
    • as the lexical component within your application
    PLIS is easily configurable, components can be extended or replaced with user generated ones to enable system customization.

    To download Java code of PLIS, go to PLIS download page at BIU NLP lab website.

  • PLIS online interactive viewer
    This is a powerful tool for investigating lexical inference processes.
    Just enter a Text and a Hypothesis, choose your knowledge resources, set a transitivity limit, choose an inference model and get the entailment probability.
    Furthermore, click on a word to see other words that can be inferred from its meaning and words that it is inferred from!

    To use the online interactive viewer, go to PLIS demo page at BIU NLP lab website.

  • WikiRules! 2.0
    A new and improved version of WikiRules! (see below): an algorithm to extract a large scale database of lexical knowledge from Wikipedia.
    New in this version:
    • You can download the database resulting in running the algorithm on the English Wikipedia dump.
    • You can download the code and run the algorithm yourself on an up to date Wikipedia dump.
    • The code now supports multiple language and can be run on Wikipedia dumps in other languages than English.

    To download the available database, go to this download page.

    To download the Java code, go to go to Wikipedia lexical miner page in GitHub of the Excitement project.

  • M-PLM: a Probabilistic Lexical Model for Ranking Textual Inferences
    A Java package of the probabilistic model for textual inference, described in Shnarch et al. (2012), which uses lexical knowledge only.
    More information and download.

  • Lexical entailment annotations
    Manually annotations of the various lexical entailment phenomena in a sample from RTE-5 pilot task.
    We have randomly sampled text-hypothesis pairs from the development set, 100 from the positive examples and another 100 from the negative. Each (T,H) pair was manually annotated for all lexical entailments between terms of the text and terms of the hypothesis.
    More information and download.

  • WikiRules!
    A large scale lexical resource automatically extracted from Wikipedia.
    The algorithm to automatically extract this rule-base from Wikipedia is described in Shnarch et. al (2009).

    To download the first version of WikiRules!, go to this download page.



  • PLIS: a Probabilistic Lexical Inference System. HTL group at FBK, Trento Italy. August 2012. slides
  • Probabilistic Lexical Models for Textual Inference. Information Retrieval group at IBM Haifa Research Lab, Israel. July 2012. slides
  • A Probabilistic Model for Lexical Entailment. ISCOL. June 2011. slides
  • Extracting a Lexical Entailment Resource from Wikipedia. IAAI Symposium. June 2008. slides