Guiding Theme C1: Motif Analysis of Text-Based Graphs

Motifs are small induced subgraphs in large networks. The motif signature of a network with respect to a selection of motifs reflects semantic characteristics of the modeled phenomena. Therefore, motif signatures are a general-purpose technique to characterize networks, to distinguish between networks with different characteristics, and to achieve a deeper understanding of the modeled phenomena. For a more systematic introduction to motif analysis and its impact on computational linguistics, read Quantifying semantics using complex network analysis by C. Biemann, S. Roos, and K. Weihe; in: Proc. COLING 2012.

A PhD project that primarily follows this guiding theme, will explore linguistic problems in Area A: Graph-based discourse processing from a structural perspective. Such a project will improve the graph-theoretic methodology but also yield a deeper understanding of the linguistic side. The student is strongly involved in the research of Area A and collaborates with the students and researchers in Area A on various linguistic questions.

Moreover, such a PhD project will combine motif analysis with methods from machine learning (ML), so the student will also collaborate with the other guiding themes of Area C: Representation and analysis of text-based structures. Motifs may be used as features in ML. Vice versa, ML may be used, for example, to find good selections of motifs or to resolve cases in which the motif signature is not sufficiently selective.

Last not least, the results of such a PhD project will be highly relevant for Area B: Language technology for multi-document summarization. Hence, the PhD student will collaborate with PhD students and researchers in Area B on identifying and achieving research goals that are useful for this area.

Poster (in German)

Example thesis topics

  • Motif analysis of linguistic networks
  • Motif analysis in multi-document summarization
  • Analysis of linguistic patterns via motifs

People

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