Speed up your customer service by automatically classifying customer tickets

A fast ticketing process is essential to provide immediate support to your customer. However, reading and classifying a large amount of tickets can be very time consuming. Deep learning can automate the ticketing  process and allow you to deal with large amounts of text very quickly, making scaling possible.

02/ The problem

Tracking the issues raised by a customer, prioritizing them and categorizing them is difficult and time consuming. If the company supports multiple channels this can be even more complicated, because a customer could raise this issue through email, phone calls, chat messages or even through tweets.

03/ The opportunity for deep learning

A deep learning model can speed up the ticketing process. The model can take a text as input and give you the ticket category as output. 
All the text data can be processed by the same model, and this can simplify ticketing for a company that supports tickets in multiple channels. Moreover, automatic classification allows you to extract valuable information from text data, making analytics possible and providing useful insights for the development of the company. If you use a multilingual model you can tackle text in multiple languages.

04/ Platform model to use

For this use case, we suggest using a text classification model, possibly multilingual.

05/ How does the model work?

The model takes a piece of text in input and learns to recognize the patterns in the text that are useful to classify the ticket in the different categories. The great thing about deep learning models is that they do not focus only on the presence or meaning of single words, but they also take the structure of the sentence and the context into consideration. This way the models have a deeper understanding of natural languages. Multilingual models can understand text on a language independent level.

06/ Data requirements

For this use case you would need around 200/300 customer tickets for each category. Your data needs to be annotated, that means that for each customer you should indicate what is the issue that they are talking about (i.e the class of the sample). For example, you can have a table with the text in one column and the category in another column. Ideally, you should have the same amount of samples for each category.

07/ Model performance and success

The evaluation view will help you understand how good your model is performing. You should check the performance of your model on each category separately. If you see that the model has a low accuracy on a particular class, this category might not be well represented in your dataset.

08/ Where to learn more

If you’re interested in building this use case, our tutorial on classifying customer complaints is a great place to start. This tutorial walks you through how to build a text classifier that can predict what kind of complaint a customer is writing about. 

You can also check out our cheat sheets in the Knowledge Center for text classification.