As the length of customer surveys shrink, clients are looking to fill the gaps with unstructured data from open-ended comments and consumer-generated media. However, processing unstructured data can be challenging and consumer-generated media are often fraught with digital exhaust - i.e. irrelevant comments.
In this talk, text analytic methods, both statistical and otherwise, are used to provide more insight for Fortune 500 customer experience programs. Several case studies will be examined using the techniques below:
- Creating rules for highly accurate comment coding.
- Identifying emerging topics using Latent Dirichlet Allocation.
- Incorporating topics and sentiments to help explain and predict CX measures such as Overall Satisfaction or Likelihood to Recommend (via Random Forests).
- Using Naïve Bayesian Classifiers to remove digital exhaust.
In this presentation, you will learn:
- Topic Modelling and How to Identify Topics via Rules in Text Analytics
- How to Determine “Key Drivers” of Overall Customer Satisfaction from comments
- How to Remove Digital Exhaust for Consumer Generated Media
Thursday, October 27, 2016
11:30 – 12:00 PM Registration & Networking & Lunch
12:00 – 1:00 PM Presentation
1:00 – 1:30 PM Q & A, Networking
$25 per attendee (Lunch included)
Cooper Irish Pub
1607 Park Place Boulevard
St. Louis Park, MN 55416
Cooper Irish Pub is located just south of 394, on the corner of 16th Street and Park Place Boulevard in the Shops at West End. Park underground year-round in the free garage beneath the pub.