Sentiment analysis application and natural language processing for mobile network operators' support on social media

Social Media have become a mixed platform of emotional expressions on services and products reviews. While network operators focus on quality of service and customer experience, mostly built upon complaints and performance indicators, subscribers and/or followers are mostly expressing their emotions on twitter and other social media. On one side, understanding followers' sentiment and perception on offered and applied services can help the South African mobile network operators to anticipate network and customer problems, embracing proactive measures rather than reactive ones on improving service and network quality. On the other side, the rise of text mining, sentiment analysis and the global Natural Language Processing expands the needs of Data Analysis across textual statistics. In this paper, we leverage on Natural Language Processing (NLP), using sentiment analysis and text mining to analyze mobile network operators, in this case CellC followers' using the R platform. We use the polarity model of sentiment analysis to determine the level of potential detraction and promotion across South African Mobile Network Operator (MNO), based on public tweets. The orientation of the study in this paper creates a bi-directional link between customers or followers and the MNO, in the goal of extracting relevant signification from sets of social media's unstructured information.