Mashine Learning Approaches to Anaphora Resolution

Student: Sukhova Ksenia

Supervisor:

Educational Programme: Fundamental and Computational Linguistics (Bachelor)

Year of Graduation: 2017

In the framework of this paper, we considered the automatic anaphora resolution without any restrictions on the anaphor (target word), considering the context data. Semantics of the context was encoded by four quantitative characteristics, computed from the word vector (word embedding, word2vec). Two types of classifiers were trained on different sets of parameters (text, vector and combined). It was found that the presence of features reflecting the semantics of the context improves the performance of the classifier. As a result, three second level models were trained on different sets of parameters. The best model, trained on combined set of parameters, reached an f-score of 46% (because of the imbalance of the data classes, the value for the minority class is "antecedent" is presented).

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