Advertisement

Informatik - Forschung und Entwicklung

, Volume 22, Issue 2, pp 71–84 | Cite as

Strategien zur webbasierten Multilingualen Fragebeantwortung

Wie Suchmaschinen zu Antwortmaschinen werden
  • Günter NeumannEmail author
Reguläre Beiträge
  • 87 Downloads

Zusammenfassung

Wir stellen eine Reihe von innovativen Methoden zur webbasierten multilingualen Fragebeantwortung in offenen Domänen vor. Insbesondere werden neuartige Strategien zur Bestimmung optimaler Antwortkontexte und zur Extraktion von exakten Antworten auf Basis von sprachunabhängigen, datengetriebenen maschinellen Lernverfahren beschrieben. Es werden zwei alternative Methoden für die crosslinguale Frageanalyse skizziert mit deren Hilfe für eine Anfrage in einer natürlichen Sprache Antworten in Dokumenten einer anderen natürlichen Sprache bestimmt werden können. Alle Methoden sind detailliert evaluiert und weisen eine vielversprechende Performanz auf.

Schlagworte

Fragebeantwortung  domänenoffen   crosslingual  datengesteuerte Methoden 

Abstract

We present a series of innovative methods for a webbased multilingual question answering in open domains. In particular, we present novel strategies for the determination of optimal answer contexts and for the extraction of exact answers on the basis of language-independent, data-driven Machine Learning algorithms. Two alternative methods for the crosslingual question analysis are presented that are used for finding answers in documents of one natural language using a query formulated in another natural language. All methods are evaluated in detail and demonstrate a promising performance.

Keywords

Question answering Open–domain Crosslingual  Data-driven methods 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

Literatur

  1. 1.
    Ankolekar A, Buitelaar P, Cimiano P, Hitzler P, Kiesel M, Krötzsch M, Lewen H, Neumann G, Sintek M, Tserendorj T, Studer R (2006) Smartweb: Mobile access to the semantic web. In: Proc of the Demo Session at the International Semantic Web Conference, Athens GA, USA, NovGoogle Scholar
  2. 2.
    Azari D, Horvitz E, Dumais S, Brill E (2003) A decision making perspective on web question answering. In: UAIGoogle Scholar
  3. 3.
    Belkin M, Goldsmith J (2002) Using eigenvectors of the bigram graph to infer morpheme identity. In: Proceedings of the ACL-02 workshop on Morphological and phonological learning, Association for Computational Linguistics, Morristown, NJ, USA, pp. 41–47Google Scholar
  4. 4.
    Clarke CLA, Cormack GV, Lynam TR, Li CM, McLearn GL (2001) Web reinforced question answering (multitest experiments for trec 2001). In: TRECGoogle Scholar
  5. 5.
    Dumais ST, Banko M, Brill E, Lin JJ, Ng AY (2002) Web question answering: is more always better? In: SIGIR, pp. 291–298Google Scholar
  6. 6.
    Echihabi A, Marcu D (2003) A noisy-channel approach to question answering. In: ACLGoogle Scholar
  7. 7.
    Figueroa A, Neumann G (2006) Genetic algorithms for data-driven web question answering. Evol Comput, accepted, to appearGoogle Scholar
  8. 8.
    Figueroa A, Neumann G (2006) Language independent answer prediction from the web. In: FinTAL, pp. 423–434Google Scholar
  9. 9.
    Holland JH (1992) Adaptation in Natural and Artificial Sys-tems. A Bradford Book. The MIT Press, Cambridge, MassachusettsGoogle Scholar
  10. 10.
    Kwok CCT, Etzioni O, Weld DS (2001) Scaling question answering to the web. In: WWWGoogle Scholar
  11. 11.
    Lita L, Carbonell J (2004) Instance-based question answering: A data-driven approach. In: EMNLP 2004, Barcelona, SpainGoogle Scholar
  12. 12.
    Magnini B, Vallin A, Ayache C, Erbach G, Penas A, de Rijke M, Rocha P, Simov K, Sutcliffe R (2005) Overview of the clef 2004 multilingual question answering track. In: Peters C, Clough PD, Jones GJF, Gonzalo J, Kluck M, Magnini B (Eds). "Multilingual Information Access for Text, Speech and Images: Results of the Fifth CLEF Evaluation Campaign. Lecture Notes in Computer Science, p. Vol. 3491Google Scholar
  13. 13.
    Moldovan D, Harabagui S, Clark C, Bowden M, Lehmann J, Williams J (2004) Experiments and analysis of lcc’s two qa systems over trec 2004. In: Proceedings of The Thirteenth Text Retrieval Conference (TREC 2004), Gaithersburg, USAGoogle Scholar
  14. 14.
    Neumann G, Piskorksi J (2002) A shallow text processing core engine. J Comput Intell 18(3):451–476CrossRefGoogle Scholar
  15. 15.
    Neumann G, Sacaleanu B (2006) Experiments on cross-linguality and question-type driven strategy selection for open-domain question answering. In: CLEF 2005, Springer-Verlag LNCS, vol 4022Google Scholar
  16. 16.
    Radev D (2003) Panel on web-based question answering. In: AAAI Spring Symposium on New Directions in Question Answering, AAAI pressGoogle Scholar
  17. 17.
    Ramakrishnan G, Paranjape D, Chakrabarti S, Bhattacharyya P (2004) Is question answering an acquired skill? In: WWW04Google Scholar
  18. 18.
    Ravichandran D, Hovy EH (2002) Learning surface text patterns for a question answering system. In: ACL, pp 41–47Google Scholar
  19. 19.
    Russell SJ, Norvig P (2004) Künstliche Intelligenz. Ein moderner Ansatz. Pearson StudiumGoogle Scholar
  20. 20.
    Sasaki Y (2005) Question answering as question-biased term extraction: a new approach toward multilingual QA. In: Proceedings of ACL’05, URL http://www.aclweb.org/anthology/P05-1027Google Scholar
  21. 21.
    Wahlster W (2000) Mobile speech-to-speech translation of spontaneous dialogs: An overview of the final verbmobil system. In: Wahlster (Ed.) Verbmobil: Foundations of Speech-to-Speech Translation, Springer, pp. 3–21Google Scholar

Copyright information

© Springer-Verlag 2007

Authors and Affiliations

  1. 1.LT–labDFKISaarbrückenDeutschland

Personalised recommendations