The semantically based computer lexicon HaGenLex. Structure and technological environment
- Sven Hartrumpf, H. Helbig, R. Osswald
- Art
- 2003
Cet article decrit les principes de conception et l'environnement technique de HaGenLex, un dictionnaire electronique general de l'allemand. Comme HaGenLex a ete developpe dans le but d'assister la…
Coreference resolution with syntactico-semantic rules and corpus statistics
- Sven Hartrumpf
- Computer ScienceConference on Computational Natural Language…
- 6 July 2001
A new hybrid approach to the coreference resolution problem is presented, which combines syntactico-semantic rules with statistics derived from an annotated corpus and is evaluated against a baseline model.
Question Answering using Sentence Parsing and Semantic Network Matching
- Sven Hartrumpf
- Computer ScienceConference and Labs of the Evaluation Forum
- 15 September 2004
The paper describes a question answering system for German called InSicht. All documents in the system are analyzed by a syntactico-semantic parser in order to represent each document sentence by a…
On metonymy recognition for geographic information retrieval
- Johannes Leveling, Sven Hartrumpf
- Computer ScienceInternational Journal of Geographical Information…
- 1 March 2008
Evaluation results indicate that removing metonymic senses from the index yields a higher mean average precision (MAP) for GIR, and a memory‐based learner is used to train a classifier and standard evaluation measures such as F‐score and accuracy are determined.
Semantic Decomposition for Question Answering
- Sven Hartrumpf
- Computer ScienceEuropean Conference on Artificial Intelligence
- 27 June 2008
This paper develops and evaluates methods for decomposing complex questions for a question answering system to less complex questions and presents six decomposition classes, which are employed for annotating the 996 different German QA@CLEF questions from 2004 till 2008 and trigger different decomposition methods.
University of Hagen at QA@CLEF 2005: Extending Knowledge and Deepening Linguistic Processing for Question Answering
- Sven Hartrumpf
- Computer ScienceConference and Labs of the Evaluation Forum
- 2005
The German question answering (QA) system InSicht participated in QA@CLEF for the second time and was improved in two main directions: the background knowledge was extended by large semantic networks and large rule sets, and coreference resolution was integrated into the parsing during document processing.
Hybrid Disambiguation of Prepositional Phrase Attachment and Interpretation
- Sven Hartrumpf
- PsychologyConference on Empirical Methods in Natural…
- 1999
The hybrid disambiguation method presented has two key features: first, it tries to solve the PP attachment problem and the PP interpretation problem, and second, it is hybrid as it combines more traditional PP interpretation rules and statistical methods.
Word Class Functions for Syntactic-Semantic Analysis
- H. Helbig, Sven Hartrumpf
- Computer Science
- 1997
Analysis with Word Class Functions (WCFA) is presented as a paradigm for syntactic-semantic analysis of natural language and a WCF machine (WCFM) which has been implemented in Lisp and is used as an NLI to bibliographic databases and multimedia databases.
A Readability Checker with Supervised Learning Using Deep Indicators
- Tim vor der Brück, Sven Hartrumpf, H. Helbig
- Computer ScienceInformatica
- 2008
Deep syntactic and semantic indicators lead to promising results comparable to surface-based indicators alone, and a graphical user interface was developed which highlights difcultpassages, depending on the individual indicator values, and displays a global readability score.
...
...