Guiding Theme B2: Content selection based on linked lexical resources

Linked lexical resources are knowledge bases that integrate different lexical resources by linking them on the sense level. In comparison to single lexical resources, they offer added value through enriched sense representations and thus their use can enhance natural language processing applications that make use of knowledge sources. For an example demonstrating the impact of linked lexical resources for the task of word sense disambiguation, read  Automated Verb Sense Labelling Based on Linked Lexical Resources by K. Cholakov, J. Eckle-Kohler and I. Gurevych, in: Proceedings of the 14th Conference of the European Chapter of the Association for Computational Linguistics (EACL 2014), p. 68-77 (PDF).

A Ph.D. project that primarily follows this guiding theme will develop methods for content selection from heterogeneous documents based on structured lexical chains. Lexical chains connect semantically related nouns, verbs and adjectives in text and are created by connecting text with the information available in lexical knowledge bases. Structured lexical chains are an extension and also take into account the results of semantic role labeling (event chains) and co-reference resolution (entity chains).

The Ph.D. project will explore the added value of linked lexical resources for the creation of structured lexical chains in collaboration with Area A: Graph-based discourse processing. This will also include the development of methods to integrate the creation of lexical chains with word sense disambiguation. Moreover, the student will closely collaborate with the guiding theme C2: Methods for contextual and constraint-based ranking on the development of ranking methods for structured lexical chains.

Poster (in German)

Example thesis topics

  • Multi-document summarization based on structured lexical chains from heterogeneous sources.
  • Parameterizable content selection based on heterogeneous sources.
  • Integration of lexical chains and discourse graphs for multi-document summarization.

References

  • Iryna Gurevych, Judith Eckle-Kohler, Silvana Hartmann, Michael Matuschek, Christian M. Meyer, and Christian Wirth: UBY – A Large-Scale Unified Lexical-Semantic Resource Based on LMF, In: Proceedings of the 13th Conference of the European Chapter of the Association for Computational Linguistics (EACL), p. 580–590, April 2012. Avignon, France.

People

A A A | Drucken Print | Impressum Impressum | Sitemap Sitemap | Suche Search | Kontakt Contact | Webseitenanalyse: Mehr Informationen
zum Seitenanfangzum Seitenanfang