SUMMARY
Virtually unknown in the business world for twenty years, text analytics has become increasingly popular. Such technique is now a crucial element of any data analytics strategy and is commonly used to analyze incident reports, maintenance reports, warranty claims, corporate reports, patents, customer feedbacks, social media, and much more. The increase applications of text analytics and the growing number of data analysts venturing into the analysis of unstructured text data, has uncover novel issues as well as common mistakes and misused by data analysts less familiar with such type of data.
In this session, we will identify the main challenges associated with the analysis of unstructured text and will describe common mistakes and misuses made by data analysts and data scientists in their attempts to extract information from text data or by managers when planning such activities. Recommendations will be made on ways to avoid the most common pitfalls of text analytics, no matter which tool you use for this.
ABOUT THE EXPERT
Normand Péladeau has a PH.D. in psychology. He is the founder of Provalis Research, a software company specialized in text analytics. He has more than 24 years of experience in the application of computer to text data and has done consulting and training for numerous multinational corporations, governments, NGOs, and academic researchers.
His current research efforts focus on improving topic modeling techniques through topic enrichments and automatic word disambiguation, as well as information extraction techniques connecting structured to unstructured data.