Churn Intent Detection in Multilingual Chatbot Conversations and Social Media

Proceedings of the 22nd Conference on Computational Natural Language Learning (CoNLL 2018), pages 161–170 Brussels, Belgium, October 31 - November 1, 2018. c⃝2018 Association for Computational Linguistics 161 Churn Intent Detection in Multilingual Chatbot Conversations and Social Media Christian Abbet†‡, Meryem M’hamdi†‡, Athanasios Giannakopoulos∗, Robert West†, Andreea Hossmann∗, Michael Baeriswyl∗ and Claudiu Musat∗ ‡ Equal Contribution ∗Data, Analytics & AI — Swisscom AG {firstName.lastName}@swisscom.com † Ecole Polytechnique F´ed´erale de Lausanne (EPFL) {firstName.lastName}@epfl.ch Abstract We propose a new method to detect when users express the intent to leave a service, also known as churn. While previous work focuses solely on social media, we show that this intent can be detected in chatbot conversations. As companies increasingly rely on chatbots, they need an overview of potentially churny users. To this end, we crowdsource and publish a dataset of churn intent expressions in chatbot interactions in German and English. We show that classifiers trained on social media data can detect the same intent in the context of chat- bots. We introduce a classification architecture that outperforms existing work on churn intent de- tection in social media. Moreover, we show that, using bilingual word embeddings, a sys- tem trained on combined English and German data outperforms monolingual approaches. As the only existing dataset is in English, we crowdsource and publish a novel dataset of German tweets. We thus underline the uni- versal aspect of the problem, as examples of churn intent in English help us identify churn in German tweets and chatbot conversations. 1 Introduction Identifying customers who intend to terminate their relation with a company is commonly known as churn detection. This is very important for companies if we consider that attracting new cus- tomers is a time and cost-intensive task. There- fore, it is often preferable for companies to focus on the existing customers in order to prevent los- ing them instead of trying to acquire new ones. Traditionally, churn detection is based on track- ing the user behavior and correlating it with the decision to churn. The analysis of the user be- havior typically includes metadata such as the sub- scription information, network usage or customer transactions (Qian et al., 2007; Dave et al., 2013). English Dictionary Churn Detection in Twitter churny non-churny Chatbot Conversations German Dictionary Mono/Multi Embeddings CNN Bi-GRU-Att Multilingual Alignment English German Figure 1: Overview of Overall Pipeline. The behavior-based techniques thus require a sig- nificant amount of data that are not easily avail- able. In addition, there is a cold start problem with novel systems which may not have access to the background required for this type of analysis. The current trend for detecting churn intent is to focus on textual user statements. This intent is sufficient evidence for the likely following churn decision of a user. Moreover, it is an action- able insight, as it allows companies to allocate re- sources to prevent the likely customer churn de- cision. Textual churn detection is only based on the current interaction between the user and the service provider. …

Content truncated to 500 words