read_me

Vaudeville Regex + Topic Model Readme 

This project aims to take vaudeville touring data form Variety Magazine using a regular expression, organize this data, and then analyze it using topic modeling. 

This task is broken down into three programs: 

The first “Vaudeville_Regex_Defense_2.2.” reads through text files and splits them into chunks based on a regular expression searching through a list of town names. 

Each of these chunks is then analyzed using a regular expression built to identify touring records. The regular expression used is intentionally capacious—incorrect results are later eliminated based on Vaudeville naming conventions. 

The program than produces a json file with the following characteristics – date (of the paper) town (town used to chunk) (venue) and (performers) 

The second program (Vaudeville_Dictionary_List_Maker_For_Topic_Modeling) takes this list, and transforms it into a flat ontology of dictionaries, each of which corresponds to a venue with a list of performers. – 

Dictionary key structure = “date,” “city_name,” “venue_name,” “artist”. The final is a list of performers. 

The third program (Touring_Topic_Model_1.py) takes this list, and does topic modelling on them using an LDA model, using every venue/performance as a date, and every performer as a word.