STRATEGIC INTELLIGENCE MODEL TO MANAGE CUSTOMER ONLINE REVIEWS THROUGH SOCIAL MEDIA

• Main Author: Rosida, Sita
• Format: Theses
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/39291
• Institution: Institut Teknologi Bandung

The role of data and information can be a valuable resource for the enterprise in the facing the uncertainty of its environmental change. Information can be regarded as a resource because of its dynamic potential value and has superior value for business in the market competition. Information that has been processed can be a knowledge that can help formulate a decision. One of these such as identifying consumer reviews through social media, the information can help top management in establishing a strategy for product or service. Assessing consumer reviews can be done by analyzing a number of relevant history data, such as consumer reviews of related features and services provided. In this research built a Model for Strategic Intelligence in Managing consumer reviews on Social Media that can assist the executive in determining marketing strategies. Model of Strategic Intelligence in managing consumer reviews on Social Media is a systematic model to analyze a number of history data to generate information achievement of one enterprise in fulfilling the ekspetasi consumers. On the implementation, this model involves a number of processes of machine learning and text mining to analyze text, i.e. text preprocessing, modeling topics, and text categorization. Topic modeling is used to assess one of the criteria of social media engagement, namely the cooperation with extraction consumer conversation information on social media, to get the consumer conversation topics. While the text categorization used to classify consumer reviews based on conversation topics in accordance the sentiment, and also to classify the status of posted company into the category of criteria, including coordination promotion, conversation, education, inspiration, and entertainment. Next text analysis results will be constructed in a quadrant of importance performance analysis for reporting on the top level management. The resulting information is expected to provide the understanding and knowledge to holistically, as well as top level management can help in making decisions for the short term tactical and strategic decisions for the long term. Experimental results show that the developed model is valid and reliable. Testing a Model of Strategic Intelligence is done by testing the validity of invalid constructs through the opinions of the expert judgements by spreading a number of questionnaires, then processed using the instrument test of validity, reliability, and measurement model. Based on the results of a test of the validity of, the value of Total Correlation for each component item in the model shows a value higher than the value of the rTabel of 0.497 and reliability test results showed the average value of the Total Cronbach's Alpha reaches 0.804. Then to test measurement model shows that the index of goodness of fit has fulfilled all the criteria for the N distribution of 100 expert judgments involved.