Hess
Fakultäten » Philosophische Fakultät » Computerlinguistik, Institut für » Prof. Dr. Michael Hess » Hess
Completed research project
| Title / Titel | Answer Extraction over Technical Manuals (ExtrAns) | ||
|---|---|---|---|
| Abstract (PDF, 14 KB) | |||
| Summary / Zusammenfassung | What is usually called ``information retrieval'' (IR) is, in actual fact, document retrieval. All IR systems return whole documents. Moreover, even the most advanced IR systems ultimately rely on keywords based approaches which means that the syntactic (and hence, the semantic) connections between the words are ignored. This means, for instance, that the strings ``computer design'' and ``design computer'' cannot be distinguished, despite the fact that they mean competely different things. Both of these problems (whole documents, lack of precision) can be taken care of if a search systems performs a (moderately deep) linguistic analysis of both queries and documents, derives a representation of (part of) their meaning, and represents it in a suitable knowledge representation language. Queries are then treated as theorems to be proved over the logical data base. That way we can retrieve those exact phrases in documents that contain the answer to the query even if they are not contiguous stretches of text. This is the basic idea of Answer Extraction. This approach requires an exact indexing scheme for individual phrases (and even words), and a colouring scheme that highlights those parts of a sentence that are most likely to contain an answer to the query. In a Swiss National Science Foundation project we are developing such as system that operates over the English language on-line manual of the Unix operating system Weitere Informationen |
||
| Publications / Publikationen | Diego Mollá Aliod, Gerold Schneider, Rolf Schwitter and Michael Hess: Answer Extraction Using a Dependency Grammar in ExtrAns. In: Traitement Automatique de Langues (T.A.L.), Special Issue on Dependency Grammar. 2000.Diego Mollá Aliod and Michael Hess: Dealing with ambiguities in an answer extraction system. In: Proc. of ATALA Workshop on Representation and Treatment of Ambiguity in Natural Language Processing. Paris. 2000.Michael Hess: Antwortextraktion über beschränkten Bereichen In: Proc. of KONVENS-98. Bonn: 1998. 337-346.Weitere Informationen | ||
| Keywords / Suchbegriffe | Natural Language Processing,, Computational Linguistics,, Information retrieval,, Document retrieval,, Answer Extraction | ||
| Project leadership and contacts / Projektleitung und Kontakte |
|
||
| Funding source(s) / Unterstützt durch |
SNF (Personen- und Projektförderung) |
||
| Duration of Project / Projektdauer | Nov 1996 to Dec 2000 |