A Framework to Analyze Social Tagging and Unstructured Data

The involvement of human interactions with business processes through Enterprise Social Networks improves organizations performance. However, Enterprise Social Networks consist of massive amount of data in form of structure and unstructured data. Therefore, finding valuable information from these types of data is a challenging issue. Nevertheless, with the annotation that are available in form of social tagging, some challenges have been resolved. In this paper, we investigate the problem of using social tagging in order to socialize organization business processes. Specifically, we present a framework to analyze social tagging and unstructured data that are generated by users to recommend tasks and activities of any type of business processes based on hybrid method of clustering and text classification. The framework uses k-means algorithm to cluster tags datasets and term frequency-inverse document frequency to weight user's documents. The experiment results performed on a real case study shows the efficiency of the framework after validates its accuracy.