Dissertations, Theses, and Capstone Projects

Date of Degree

9-2017

Document Type

Thesis

Degree Name

M.A.

Program

Linguistics

Advisor

Martin Chodorow

Subject Categories

Anthropological Linguistics and Sociolinguistics | Computational Linguistics

Keywords

sociolinguistics, sentiment analysis, word embedding models, topic modeling

Abstract

Since Robin Lakoff started the conversation around language and gender with her 1975 essay “Language and Woman’s Place,” extensive work has been done on analyzing sociolinguistics associated with gender. While much work has been done on the differences between how men and women use language, there is less research to be found on language about women as opposed to language about men. In this work, I build a word embedding model from a corpus of Wikipedia film summaries and use this model to create lists of words associated with men and words associated with women. I then use sentiment analysis tools to assess the emotional valence of these words and sentences containing them. I find that when comparing words and sentences associated with men and women, language about women tends to be more consistently more positively valenced, while words associated with men cover a wider breadth of valences, skewing slightly negative.

Share

COinS