Purpose: It is said that newspaper is something that we cannot get rid of. It is very necessary for us to know about our surroundings and to know about the things in our world. In today’s situation, people like to read everything online, unlike newspaper printed on paper which is commonly found in today's technological world. For the reader, it is very good to save time to get each news item separately. But readers have made everything online and the biggest problem is in the area of reading newspapers once they get all the news together without separating them. Not only that, but in our newspaper industry, sorting the news into each category at the time of printing requires a lot of labor. Therefore, many research works has been done in various languages regarding the categorization of news. This paper discusses a detailed literature survey of various approaches used to categorize news headlines in Malyalam language. Design/Methodology/Approach: The details collected for this review paper were obtained by analysing and comparing different research articles from recognized resources. Objective: To find a research gap and appropriate solutions for news headlines categorizations in Malayalam. Results/ Findings: Review of this paper gives a proper understanding of news headlines categorization   in Malayalam and other languages in connection with machine-learning and deep learning approaches. Originality/Value: The review of this paper exhibits an analysis of machine learning algorithms for news headlines categorization in Malayalam and other languages and suggests the importance of news headlines categorization.