Unsupervised Relation Extraction from Russian News Texts

Student: Nefedov Mikhail

Educational Programme: Language Theory and Computational Linguistics (Master)

Year of Graduation: 2017

In this paper we describe methods for unsupervised relation extraction. We test three popular methods on russian news texts: bootstrapping, distant supervision and relation extraction with topic modelling. Based on of the results we conclude that bootstrapping is better for relations that are expressed uniformly in texts, whereas distant supervision is better when the combination of named entities is not ambigous. We fail to successfuly apply topic modeling for relation extraction. Potential reasons for that are insufficient dataset size and errors in named entity recognizer and dependency parser.

Student Theses at HSE must be completed in accordance with the University Rules and regulations specified by each educational programme.

Summaries of all theses must be published and made freely available on the HSE website.

The full text of a thesis can be published in open access on the HSE website only if the authoring student (copyright holder) agrees, or, if the thesis was written by a team of students, if all the co-authors (copyright holders) agree. After a thesis is published on the HSE website, it obtains the status of an online publication.

Student theses are objects of copyright and their use is subject to limitations in accordance with the Russian Federation’s law on intellectual property.

In the event that a thesis is quoted or otherwise used, reference to the author’s name and the source of quotation is required.

Search all student theses