Clustering online customer complaints
International Journal of Business Quantitative Economics and Applied Management Research, vol.3, no.7, pp.1-18, 2016 (Peer-Reviewed Journal)
- Publication Type: Article / Article
- Volume: 3 Issue: 7
- Publication Date: 2016
- Journal Name: International Journal of Business Quantitative Economics and Applied Management Research
- Journal Indexes: Other Indexes
- Page Numbers: pp.1-18
- Open Archive Collection: AVESIS Open Access Collection
- Erciyes University Affiliated: Yes
Abstract
Nowadays one of the biggest needs of a firm is to extract knowledge by analysing unstructured big data in the strategic decision making process in order to improve customer satisfaction. The aim of this research is automatically clustering the online complaints which are about the customers ignoring their subscriptions in order to understand a specific group of the complaints. 809 customer complaints which are about ignoring subscriptions are collected from huge amount of online complaints with web mining from a telecommunication firm and those of its biggest competitor in Turkey. Text mining and natural language processing techniques are used to analyse the data. The positive feedback of the customers are re-analysed to make an adaptation. We present an adaptive feedback model to achieve knowledge which helps in making business decisions. We clustered online complaints while determining similarities between the groups. New, interesting and hidden knowledge about customer complaints are found. Index Terms— Web text mining, Customer feedback, Natural language processing, Clustering