Turn Email into Data with Deep Learning, Plus other Industry Tasks with Gensim Topic Modeling
I will tell of two business problems that can be solved using open source NLP tools. They are stories of two real consulting projects at RaRe Technologies. The business question of "how is the celebrity content monetising this month" was solved with a Topic Modelling algorithm from a popular open-source NLP package Gensim that we maintain. It gave a birds eye view of the entire content library of a mass media company client. I will also talk about how to extract information from unstructured text into a structured table. The example project here will be a deep learning model parsing emails on about logging truck movements across Canada. The text came in 100 different formats that were constantly changing but the deep network could parse even the unseen formats.