Clustering Methods for Multidimensional Data from Social Media

• Published in: 2024 MIT Art, Design and Technology School of Computing International Conference (MITADTSoCiCon) pp. 1 - 7
• Main Authors: Ikramuddin, Ikramuddin, Avasthi, Sandhya, Tyagi, Meenakshi
• Format: Conference Proceeding
• Language: English
• Published: IEEE 25.04.2024
• Subjects: clustering; Clustering algorithms; consensus function; Ensemble clustering; Learning systems; Machine learning; Machine learning algorithms; Reviews; Social networking (online); unlabeled data; unsupervised learning methods; Video on demand
• DOI: 10.1109/MITADTSoCiCon60330.2024.10575244

Popular platforms like Instagram, Facebook, YouTube, Linkedin, and Twitter have become essential tools for various purposes generating massive amounts of unstructured data. Fast processing and analysis of this data need efficient Machine Learning methods. Clustering algorithms can organize and group such data efficiently and that's why efficient clustering algorithms are need of the hour. The data coming through social media platforms has various dimensions and a single algorithm cannot extract all the dimensions. Ensemble clustering is an efficient model by aggregating multiple base clustering algorithms that deal with the same data. This review paper explores different methods of ensemble clustering and performs the comparison of the same. In addition, discusses different ways to make consensus function efficient to explore insights from the given multidimensional data and how different strategies can be used in building an efficient model for clustering. But still, none of the clustering or ensemble clustering models or algorithms can be said this be the ideal one. Because as the need arises different clustering algorithm is used. So keep in mind that whether it is clustering or ensemble clustering if we apply the two different algorithms to the same data the result is different and as per the need the clustering models and algorithms are used.