Automatic Topic Classification And Document Clustering Using Lda Based Machine Learning Techniques

Since more and more information is being stored in digital formats (such as databases, journals, libraries, and papers), it is crucial that this information be put to good use. However, it might be challenging to extract meaningful insights from the data because of its poor organisation, ambiguity, and lack of precision. Data may be correctly categorised to allow for separate research on different types of data, which will help with this problem. Because of this, researchers in disciplines like mathematics and computer science have been trying to figure out how to fix the issue. Material mining is a branch of machine learning that helps cope with unstructured and disorganised data by sifting through it for useful insights. Text mining is a subfield of data mining that focuses on discovering patterns and extracting structured attributes from unstructured text. To successfully analyse such data, supervised learning is combined with unsupervised learning. It has become common practise to employ Latent Dirichlet Allocation (LDA), a probabilistic approach to topic modelling, when classifying documents. As an unsupervised method, however, traditional LDA's findings may shift depending on which articles come first in the analysis. In this study, we provide a comprehensive strategy for avoiding the order effect as well as the unsupervised probabilistic character of the problem. The proposed framework achieved a success rate of 95.56% & 95.23% when evaluated on two separate datasets, demonstrating its better classification capabilities. Furthermore, the subjects that were extracted from the classified clusters are highly relevant to the domain classifications.