Hybrid News Recommendation System using TF-IDF and Similarity Weight Index
C.P. Patidar1, Yogesh Katara2, Meena Sharma3
1C. P. Patidar*, Assistant Professor, Department of Information Technology, IET DAVV, Indore, India.
2Yogesh Katara*, PG Student, Department of Information Technology, IET DAVV, Indore, India.
3Dr. Meena Sharma, Professor, Department of Computer Science, IET DAVV, Indore, India.
Manuscript received on October 02, 2020. | Revised Manuscript received on October 05, 2020. | Manuscript published on November 30, 2020. | PP: 5-9 | Volume-10 Issue-3, November 2020. | Retrieval Number: 100.1/ijsce.C34711110320 | DOI: 10.35940/ijsce.C3471.1110320
Open Access | Ethics and Policies | Cite | Mendeley
© The Authors. Published By: Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: As the usage of internet is increasing, we are getting more dependent on it in our daily life. The Internet plays an essential role to simplify our tight schedules. In such tough lives, it is very important to stay aware of current affairs. Now for different people coming from different backgrounds and professions, the preferences are different too. Here come Data mining techniques in the picture, which gives us “Recommender system” as the output, capable of delivering more relevant and worthy outcomes. Newspapers are the basic obligation asked by almost every person to stay updated and aware of the world. But as we observe that nowadays, various solutions are been developed to convert paper news system to digital news and raise the bar of the quick news. And that’s how News Recommender systems are have made an important place in our fast running lives.This research paper has investigated the News Recommendation solution right from its core, including the importance, performance, and improvement suggestions. This paper talks about enhancing the performance of states solution by using modified Term Frequency-Inverse Document Frequency (TF-IDF) algorithms. Proposed solution advocates the usage of JAVA technology which reflects fruitful results in the final graphs of accuracy, precision, and F-score. Here, BBC dataset has been used for comparison study purposes.
Keywords: Associative Calculus, BBC Dataset, News Recommendation, TF-IDF.