Interlinking Entities in Newspaper Archives for enhanced Searchability and Depiction

Thesis Type Master
Thesis Status
Student Andreas Schmid
Thesis Supervisor

The goal of this master thesis was to enable for searching names, locations and organisations in the digitized archives of a local newspaper, and to categorize these. Full text searching is possible as well as searching for named entities or their categories, which are taken from Wikipedia. This later allows to search and find on one hand articles mentioning a specific person but also whole topics like articles about persons who were representatives in the Austrian parliament in 1990.

Additionally, an effort is taken to evaluate building a manually annotated corpus directly from the newspaper instead of using an already existing one, which is built from the German newspaper "Die Tageszeitung".