Using Sentiment Analysis To Fill In The Gaps From User Surveys
We put a year's worth of online help chat logs from a major Australian Superannuation website through Google's Natural Language API, to see what insights we could gain from the users. This talk will discuss how the Natural Language API works, and the underlying machine learning concepts, and also give you some ideas on how to make use of the information based on examples from our work. We'll compare the sentiment values with those expressed in exit surveys and find out how useful an indicator it can be.
Outline/Structure of the Case Study
I'll start by describing the business problem we were solving, how we set up the analysis, how the natural language API works and finally share some of the insights we gained.
Learning Outcome
Participants will leave knowing more about sentiment analysis, and how to apply it to their data.
Target Audience
Developers that want to know how sentiment analysis works, data analysts that want to squeeze more information from chat logs.
Prerequisites for Attendees
No prior knowledge expected or required.
Video
Links
I spoke at YOW! Data last year on Image Recognition: https://www.youtube.com/watch?v=QlsEurv4xa8
I've also spoken at YOW! Connected, and Melbourne's Google DevFest.
Dave Thomas
Josh Graham
Grant Carlson
Brian Thorne
James Binh Tham
Public Feedback