Deductive Text Analytics: Using Categorization Models for Confirmatory Approaches and Hypothesis Testing by Dr Derrick L. Cogburn
In this session, Dr. Cogburn will introduce participants to categorization models, a powerful deductive technique for text mining. Researchers build categorization models (also known as dictionaries or lexicons), to represent a specific concept or construct of interest. These approaches help to identify the extent to which a specific concept is present or absent within a specific large-scale textual dataset. These categorization models can be relatively simple bifurcations of concepts, or require multiple levels and sub-levels supported by specific words, phrases and/or rules to help identify the specific content of interest. There are many intriguing applications of this technique, one of the more popular being sentiment analysis, where text may be categorized as containing “positive” or “negative” sentiment. Dr. Cogburn will discuss how to either identify and use existing categorization models, or to build, validate and deploy your own models to answer specific research questions or test hypotheses.