Ticket Sentiment App

Install this app through Zendesk Marketplace.

What is sentiment score? 

There are many methodologies to calculate sentiment scores but overall, the science is based on algorithms of natural language processing as well as machine learning to categorize pieces of words as positive, neutral, or negative. This app is using AFINN lexicon. Thanks to Finn Årup Nielsen for compiling list of 3,300+ words with polarity score associated with each word. When a ticket requester says the word "happy" for example, this adds 3 points to the score. At the same time, if the customer says the word "does not work", this deducts 3 points to the overall score. You can see the complete list of words on this page

What are the limitations of this app? 

This app currently only works with English. There are plans to incorporate other languages at future date (Please let us know! We'd love to hear your feedback). As with other machine learning technologies, this app has flaws when compared to other app using different methodologies. This app may not be 100% accurate and may not work with recent jargons or acronyms.

What's coming? 

We're working on other languages (e.g. French, Polish, Swedish, Finnish, Danish, and Turkish).