Deep analysis in IQA: evaluation on real users dialogues.
Název práce v češtině: | |
---|---|
Název v anglickém jazyce: | Deep analysis in IQA: evaluation on real users dialogues. |
Akademický rok vypsání: | 2008/2009 |
Typ práce: | diplomová práce |
Jazyk práce: | angličtina |
Ústav: | Ústav formální a aplikované lingvistiky (32-UFAL) |
Vedoucí / školitel: | doc. RNDr. Vladislav Kuboň, Ph.D. |
Řešitel: | skrytý - zadáno a potvrzeno stud. odd. |
Datum přihlášení: | 17.11.2008 |
Datum zadání: | 21.11.2008 |
Datum a čas obhajoby: | 14.09.2009 00:00 |
Datum odevzdání elektronické podoby: | 14.09.2009 |
Datum proběhlé obhajoby: | 14.09.2009 |
Oponenti: | PhDr. Petra Hoffmannová |
Zásady pro vypracování |
Question Answering (QA) systems have reached a high level of
performance within the scenario originally described in the TREC competitions, and are ready to tackle new challenges as shown by the new tracks proposed in recent instantiations [4]. To answer these challenges, attention is moving towards adding semantic information at different levels. Our work is about context modeling for Interactive Question Answering (IQA) systems [2,3]. We will take as a case study "BoB", the chatterbot developed at Free University of Bozen-Bolzano. The current system use a statistical approach based on regression models. [1] The task of the thesis is to conduct evaluation experiments with data collected by dialogues with real users via the on-line beta-version of BoB [5]. Based on the result of this analysis, the thesis should propose suggestions of formal semantics and deep analysis modules to be integrated in the current system architecture so to overcome the identified limitations. In particular, attention will be on the role of anaphora, ellipses, as well as the need of background knowledge for understanding Follow up questions. |
Seznam odborné literatury |
[1] Raffaella Bernardi and Manuel Kirschner Context Modeling for IQA:
The Role of Tasks and Entities In Proceedings of KRAQ. [2] Chai, Joyce Y. and Rong Jin. 2004. Discourse structure for context question answering. In Proc. of the HLT- NAACL 2004 Workshop on Pragmatics in Question Answering, Boston, MA. [3] van Schooten, Boris, R. op den Akker, R. Rosset, O. Galibert, A. Max, and G. Illouz. forthcoming. Follow-up question handling in the IMIX and Ritel systems: A comparative study. Journal of Natural Language Engineering. [4] Voorhees, Ellen M. 2004. Overview of the TREC 2004 question answering track. In Proc. of the 13th Text REtrieval Conference. [5] BoB: http://www.unibz.it/library/askbob/ |
Předběžná náplň práce |
The task of the thesis is to conduct evaluation experiments with data collected by dialogues with
real users via the on-line beta-version of BoB [5]. Based on the result of this analysis, the thesis should propose suggestions of formal semantics and deep analysis modules to be integrated in the current system architecture so to overcome the identified limitations. In particular, attention will be on the role of anaphora, ellipses, as well as the need of background knowledge for understanding Follow up questions. |
Předběžná náplň práce v anglickém jazyce |
The task of the thesis is to conduct evaluation experiments with data collected by dialogues with
real users via the on-line beta-version of BoB [5]. Based on the result of this analysis, the thesis should propose suggestions of formal semantics and deep analysis modules to be integrated in the current system architecture so to overcome the identified limitations. In particular, attention will be on the role of anaphora, ellipses, as well as the need of background knowledge for understanding Follow up questions. |