A framework for streamlined statistical prediction using topic models
This provides a streamlined framework for researchers in Humanities and Social Sciences to incorporate NLP techniques into traditional statistical methods, but it is incremental as it adapts existing tools.
The paper tackles the problem of integrating topic models into classical statistical prediction frameworks for text in Humanities and Social Sciences, showing that topic regression models perform comparably to less efficient word-based predictors.
In the Humanities and Social Sciences, there is increasing interest in approaches to information extraction, prediction, intelligent linkage, and dimension reduction applicable to large text corpora. With approaches in these fields being grounded in traditional statistical techniques, the need arises for frameworks whereby advanced NLP techniques such as topic modelling may be incorporated within classical methodologies. This paper provides a classical, supervised, statistical learning framework for prediction from text, using topic models as a data reduction method and the topics themselves as predictors, alongside typical statistical tools for predictive modelling. We apply this framework in a Social Sciences context (applied animal behaviour) as well as a Humanities context (narrative analysis) as examples of this framework. The results show that topic regression models perform comparably to their much less efficient equivalents that use individual words as predictors.