Effectively using Semantic Similarity Learning for Mining Hidden Social Network Contents

---

Abstract

Individuals and organizations are increasingly relying on social media to communicate. Massive volumes of publicly accessible data are stored on social media platforms, making it a great source of knowledge and insight. Text mining may be useful for generating insights from language data; however, it can be difficult to effectively deduce sense using social media text based on a single social media account. The study presents a technique for mining brief text structures to deduce the user's overarching themes from commonly appearing terms in social media accounts. The cosine textual similarity approach is used to determine the degree of similarity between two texts. It uses a clustering label propagation approach for labeling the text. This approach may be beneficial for getting decision-making insights from social media or other online forms that include short or sparse language.

---

Keywords

Clustering; Machine Learning (ML); Social Network Analysis (SNA); Mining.