Abstract—Using Apache SparkTM, Natural Language Processing (NLP), and text mining, the author’s team analyzed enterprise Customer Relationship Management (CRM) data to predict and combat premature sales attrition. NLP techniques were employed to identify critical patterns in unstructured CRM data fields. The Apache Spark big data engine and the DatabricksTM big data software as a service (SaaS) platform were then used to efficiently locate evidence of these patterns within large amounts of unstructured CRM data. New business reports are being generated from these patterns so that in-house support staff may proactively engage with customers. Future work will focus on automating these techniques to identify new problems and to quickly spot emerging trends.

Index Terms—big data, apache spark, Natural Language Processing (NLP), Customer Relationship Management (CRM) software

Cite: Keith Gutfreund, "Big Data Techniques for Predictive Business Intelligence" Journal of Advanced Management Science, Vol. 5, No. 2, pp. 158-163, March 2017. doi: 10.18178/joams.5.2.158-163