Influence of language morphological complexity on information retrieval
Author(s)
Dolamic, Ljiljana
Date issued
2010
Subjects
stemmer natural language processing with Indo-European languages search engines with asian languages cross-language information retrieval (CLIR) automatic translation
Abstract
In this dissertation two aspects of information retrieval are elaborated. The frst involves the creation and evaluation of various linguistic tools for languages less studied than English, and in our case we have chosen to work with the two Slavic languages Czech and Russian, and three languages widely spoken on the Indian subcontinent, Hindi, Marathi and Bengali. To do so we compare various indexing strategies and IR models most likely to obtain the best possible performance. The second part involves an evaluation of the effectiveness of queries written in different languages when searching collections written in either English or French. To cross the language barriers we apply publicly available machine translation services, analyze the results and then explain the poor performances obtained by the translated queries.
Notes
Thèse de doctorat : Université de Neuchâtel, 2010
Publication type
doctoral thesis
File(s)