Clustering Documents based on Semantic Similarity using HAC and K-Mean Algorithms

The continuing success of the Internet has greatly increased the number of text documents in electronic formats. The techniques for grouping these documents into meaningful collections have become mission-critical. The traditional method of compiling documents based on statistical features and grouping did use syntactic rather than semantic. This article introduces a new method for grouping documents based on semantic similarity. This process is accomplished by identifying document summaries from Wikipedia and IMDB datasets, then deriving them using the NLTK dictionary. A vector space afterward is modeled with TFIDF, and the clustering is performed using the HAC and K-mean algorithms. The results are compared and visualized as an interactive webpage.