October 31, 2019 Joris van Gend

Natural Language Processing (NLP)

Recent advances in both computer hard- and software have allowed for evermore complex analysis. In the same period the ability analyze texts as a human would do has increased immensely. Big players like Google, Amazon and IBM have invested big recourses in both the reading analysis of texts. As a result computers are able to attach meaning, link concepts and process raw conversational data. In the domain of organizational research this allows for much faster and reliable coding and analysis of far larger amounts of conversational data.

Analyzing texts with IBM Watson and Python

The BMS lab offers you several methods to analyze speech like text. Both hardware and software solutions exist. They range in complexity and ability from our very own Sociometry app to the deep learning capabilities of IBM’s Watson. The platform of IBM is fed and improved with all the data it is provided with an is able to understand concepts in both text and data files. For example, if Watson is fed a datafile with the salesdata of the past 6 months of a company, it can understand many of the concepts such as coupons, turnover and profit and present automatic conclusions. It does the same with text files and this well trained and honed deep learning network can provide you with truly unique perspectives. The BMS lab has licenses available for this software. For those not afraid of a bit of computer code, there are Python suites like Anaconda, wherein fairly advanced text analysis can be conducted.