trensanton/detecting-the-conceptual-evolution-of-social-inequality-in-sociology

Detecting-the-Conceptual-Evolution-of-Social-Inequality-in-Sociology

Social inequality regarding race, gender, class, and other social identities is a key issue in oursociety as well as a crucial topic in many fields of social studies. We aim to explore how the conceptual discussion, analysis, and criticism of social inequality problems evolve in sociology papers. However, because there are numerous top journal papers whichare often very lengthy (over 30 pages), it is difficult and time-consuming to explore all papers byhuman reading. In view of this difficulty, we need computational models to explore them.

This project aims to modify and adapt text mining and machine learning methods, previously developed based on conventional computerscience tasks and benchmark datasets, to detect the conceptual evolution of social inequality inprestigious sociology journals. Therefore, our focus will be on new ways of using and evaluatingtopic modeling and word embedding for the purpose of analysis. It is worth noting that, comparedto pure substantial analysis, we are more interested in the application and evaluation of machinelearning models in social data mining for analysis. This paper, accordingly, will focus more onthe technical aspects in terms of application and evaluation for analysis rather than substantialanalysis of conceptual evolution itself.