An Improved Movie Recommendation System Using Advanced Recommendation Techniques

Abstract

We live in a world where people really need something to do. People need entertainment to stay happy and get things done. Watching movies is a way for people to relax and feel good. Movies are a way to unwind and feel better. There are many movies on the internet. It can be really hard to find a people find movies they will really enjoy. The system is for people who watch movies. Movies are what the system is, about. This System is called as Hybrid Recommendation System. It uses two ways of finding movies: one that looks at the movie itself and one that looks at what other people like you have watched. The system puts together what people do when they watch movies and what the movies are about to make suggestions. We also use a tool to make sure the suggestions are accurate. We use another tool to pick the best features of the movies so the system works well.

Introduction

The text presents a Hybrid Movie Recommendation System designed to improve the accuracy, quality, and scalability of movie recommendations by combining content-based filtering, collaborative filtering, clustering, Support Vector Machines (SVM), genetic algorithms, and cosine similarity.

1. Background and motivation

With the rapid growth of online movie databases and the increasing number of users, finding suitable movies manually has become difficult. Movie recommendation systems help users discover films based on preferences such as genre, cast, director, language, release year, and user ratings.

However, traditional recommendation approaches face problems such as:

• Low recommendation quality.

• Difficulty handling large numbers of users and movies.

• Scalability problems.

• Limitations of using only content-based or collaborative filtering.

The study therefore proposes a hybrid approach intended to improve both recommendation quality and computational efficiency.

2. Related work

Previous research explored several techniques:

• Fuzzy C-means, weighted similarity, genetic algorithms, and clustering were combined to improve movie similarity and recommendation quality, although computational time increased.

• MovieGEN used machine learning, clustering, and Support Vector Machines to narrow down movies based on user responses and generate recommendations.

These studies motivate the use of a combination of machine-learning and optimization techniques.

3. Proposed techniques

Content-Based Filtering

The system analyzes movie characteristics and recommends movies similar to those a user has previously preferred. Its main stages are:

1. Term allocation.

2. Term representation.

3. Selection of a learning algorithm.

4. Generation of recommendations.

Collaborative Filtering

The system also considers relationships between users and their movie preferences. Two algorithms are discussed:

• SVM: Used as a classification/prediction technique.

• Adjusted K-Means: Groups similar movies or users into clusters by repeatedly assigning objects to the closest cluster and updating cluster centers.

Genetic Algorithm

A Genetic Algorithm (GA) is used to optimize the recommendation process. It follows steps such as:

• Population initialization.

• Fitness evaluation.

• Selection of the best solutions.

• Crossover.

• Mutation.

• Creation of a new population.

• Repetition until a termination condition is reached.

The goal is to identify parameter combinations that produce better recommendations.

Cosine Similarity

Cosine similarity measures how similar two movie-related vectors are by calculating the angle between them. It focuses on the orientation of vectors rather than their magnitude, making it useful for comparing movie feature representations.

4. Dataset

The study uses three different MovieLens-related datasets of different sizes:

• OMDB Lens 5000

• MovieLens Small Latest

• TMDB Lens 10M

The datasets contain movie ratings on a 1–5 scale, along with user information in some datasets.

Using datasets of different sizes allows the researchers to examine whether the proposed hybrid system remains effective as the amount of data increases.

5. Evaluation metrics

The system is evaluated using several standard recommendation metrics:

• Mean Absolute Error (MAE): Measures the difference between predicted and actual user ratings. Lower MAE indicates better prediction.

• Precision: Measures how many recommended movies are actually relevant to the user.

• Recall: Measures how many relevant movies the system successfully recommends.

• F-measure: Combines precision and recall into a single measure.

• Coverage: Measures the proportion of movies for which the system can generate recommendations.

The study also evaluates:

• Accuracy

• Scalability

• Computational time

6. Main findings

According to the reported experimental results, the hybrid recommendation system performs better than standalone content-based or collaborative filtering approaches. The combination of methods is reported to provide:

• More accurate recommendations.

• Better recommendation quality.

• Fewer prediction errors.

• Greater ability to handle increasing numbers of users and movies.

• Improved computational efficiency.

The Genetic Algorithm further improves the system by optimizing relevant parameters and helping identify better recommendation configurations.

Conclusion

In this paper, to improve the accuracy, quality and scalability of movie recommendation system, a Hybrid approach by unifying content based filtering and collaborative filtering; using Support Vector Machine as a classifier and genetic algorithm is presented in the proposed methodology. Existing pure approaches and proposed hybrid approach is implemented on three different Movie lens datasets and the results are compared among them. Comparative results depicts that the proposed approach shows an improvement in the accuracy, quality and scalability of the movie recommendation system than the pure approaches. Also, computing time of the proposed approach is lesser than the other two pure approaches.

References

[1] Hirdesh Shivhare, Anshul Gupta and Shalki Sharma (2015), “Recommender system using fuzzy c- means clustering and genetic algorithm based weighted similarity measure”, IEEE International Conference on Computer, Communication and Control. [2] Manoj Kumar, D.K. Yadav, Ankur Singh and Vijay Kr. Gupta (2015), “A Movie Recommender System: MOVREC”, International Journal of Computer Applications (0975 – 8887) Volume 124 – No.3. [3] RyuRi Kim, Ye Jeong Kwak, HyeonJeong Mo, Mucheol Kim, Seungmin Rho,Ka Lok Man, Woon Kian Chong (2015), “Trustworthy Movie Recommender System with Correct Assessment and Emotion Evaluation”, Proceedings of the International MultiConference of Engineers and Computer Scientists Vol II. [4] Zan Wang, Xue Yu*, Nan Feng, Zhenhua Wang (2014), “An Improved Collaborative Movie Recommendation System Using Computational Intelligence”,Journal of Visual Languages & Computing,Volume 25, Issue 6. [5] Debadrita Roy, Arnab Kundu, (2013), “Design of Movie Recommendation System by Means of Collaborative Filtering”, International Journal of Emerging Technology and Advanced Engineering, Volume 3, Issue 4. [6] Sajal Halder, A. M. Jehad Sarkar, Young-Koo Lee (2012), “Movie Recommendation System Based on Movie Swarm”, Second International Conference on Cloud and Green Computing. [7] Birtolo, C., Ronca, D., Armenise, R., &Ascione, M. (2011), “Personalized suggestions by means of collaborative filtering: A comparison of two different model- based techniques”, In NaBIC, IEEE (pp. 444–450). [8] Jesus Bobadilla, Fernando Ortega, Antonio Hernando and Javier Alcala (2011) “Improving collaborative filtering recommender system results and performance using genetic algorithms”, ELSEVIER Knowledge Based Systems, Vol. 24, issue 8, pp. 1310–1316. [9] Mustansar Ali Ghazanfar and Adam Prugel-Bennett (2010), “Building Switching Hybrid Recommender System Using Machine Learning Classifiers and Collaborative Filtering”, IAENG International Journal of Computer Science, 37:3, IJCS_37_3_09. [10] J. Bobadilla, F. Serradilla and J. Bernal (2010) “A new collaborative filtering metric that improves the behavior of recommender systems”, ELSEVIER Knowledge-Based Systems, Vol. 23, Issue 6, pp. 520-528. [11] J.L. Sanchez, F. Serradilla, E. Martinez, J. Bobadilla, (2008) “Choice of metrics used in collaborative filtering and their impact on recommender systems”, in: Proceedings of the IEEE International Conference on Digital Ecosystems and Technologies DEST, pp. 432– 436. [12] Eyrun A. Eyjolfsdottir, Gaurangi Tilak, Nan Li (2008), “MovieGEN: A Movie Recommendation System”, 2008 Conference Proceedings. [13] J. Ben Schafer, ShiladSen, Dan Frankowski, Jon Herlocker, (2007) “Collaborative filtering recommender systems” In the Adaptive Web, pp. 291–324. Springer Berlin / Heidelberg. [14] F. Kong, X. Sun, S. ye, (2005), “A comparison of several algorithms for collaborative filtering in startup stage”, in: Proceedings of the IEEE Networking, Sensing and Control, pp. 25–28. [15] S.V.N. Vishwanathan, M. Narasimha Murty (2002), “SSVM: A Simple SVM Algorithm”, Neural Networks, IJCNN \'02. Proceedings of the International Joint Conference on (Volume: 3) [16] Joseph Konstan, George Karypis, BadrulSarwar, and John Riedl, (2001) “Item-Based Collaborative Filtering Recommendation Algorithms”, ACM.

Copyright

Copyright © 2026 Shesh Mani Tiwari, Mohd. Shah Alam, Ajay Tiwari, Amit Katiyar, Abhishek Dwivedi. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.