Назаров Олег Витальевич
Мониторинг клиентского опыта на базе 360° customer view: модели и структура
Системы больших данных
360-degree customer view approach in business is overwhelming. The main idea of this approach is development single and integrated customer view using all available information about customer – structured and unstructured (web logs, social network and etc.) data. The research describes different modern functional architectures, which support 360-degree customer view. In addition, this research proposes a completed developing model technique that improves churn modeling though this approach, which includes text mining and network analysis processing. The sample of customers of one most popular bank in Russia, which represents a real case of customer relationship management (CRM), is used. Textual data attributes of customers have been downloaded from VK and Facebook (popular social networks in Russia). The churn model based on the text mining and network analysis pre-processing and linear models (for example, logistic regression) is proposed and evaluated through a comparison with standard model, which uses only internal attributes from CRM database. The results show that adding the attributes, which are developed by text mining and network analysis, allows to improve classification accuracy.