Recommender System In Youtube Based On Sentiment Based Model
Abstract
In the era of personalized digital experiences,
recommender systems play a pivotal role in enhancing
user engagement. Traditional recommendation
techniques heavily rely on user ratings, which often
suffer from limitations such as the cold start problem,
leading to inaccurate suggestions when user data is
sparse or inconsistent. To address this challenge, this
project proposes a sentiment-based model that
incorporates user comments alongside ratings to
improve recommendation accuracy. By analyzing
sentiments expressed in comments, deeper insights into
user preferences are obtained, enabling more reliable
predictions. A CNN2D (Convolutional Neural Network
2D) model is trained on YouTube comment data
categorized into five sentiment levels, from negative to
extremely happy, allowing a nuanced interpretation of
user feedback. The model's performance is evaluated
using RMSE (Root Mean Square Error) metrics,
indicating prediction accuracy with dynamically split
training and testing datasets. The application, built with
Python and MySQL, offers functionalities such as user
registration, dataset loading, model training, and both
file-based and single-comment analysis for sentiment
prediction and movie recommendation. With a userfriendly
interface and detailed modules, the system
allows seamless interaction, accurate sentiment
classification, and effective recommendation
generation. This approach not only mitigates the cold
start issue but also enriches the recommendation
quality by leveraging the emotional context found in
user comments.
Downloads
References
Pang, B., Lee, L., & Vaithyanathan, S. (2002).
Thumbs up? Sentiment classification using
machine learning techniques. Proceedings of the
ACL-02 conference on Empirical methods in natural
language processing, 79-86.
2. Liu, B. (2012). Sentiment analysis and opinion
mining. Synthesis Lectures on Human Language
Technologies, 5(1), 1-167.
3. McAuley, J., & Leskovec, J. (2013). Hidden factors
and hidden topics: Understanding rating
dimensions with review text. Proceedings of the 7th
ACM conference on Recommender systems, 165–
172.
4. Kim, Y. (2014). Convolutional neural networks for
sentence classification. arXiv preprint
arXiv:1408.5882.
5. Tang, J., Hu, X., & Liu, H. (2015). DeepCoNN:
Deep cooperative neural networks for
recommendation. Proceedings of the 2015 ACM on
International Conference on Information and
Knowledge Management, 305–314.
6. Covington, P., Adams, J., & Sargin, E. (2016). Deep
neural networks for YouTube recommendations.
Proceedings of the 10th ACM conference on
Recommender systems, 191–198.
7. Zhang, Y., & Lai, G. (2017). Sentiment-aware
recommendation based on hybrid deep learning.
Proceedings of the 26th International Joint
Conference on Artificial Intelligence (IJCAI), 3945-
3951.
8. Chen, C., Zhang, M., Ma, W., Ma, S., Liu, Y., & Ma,
J. (2018). Neural sentiment recommendation.
Proceedings of the 41st International ACM SIGIR
Conference on Research and Development in
Information Retrieval, 123–132.
9. Musto, C., Semeraro, G., Lops, P., & de Gemmis, M.
(2018). VSM-based context-aware
recommendation exploiting opinion mining
techniques. Information Systems, 72, 1-15.
10. Wu, L., Sun, P., Fu, Y., Hong, R., & Wang, X.
(2019). A neural influence diffusion model for
social recommendation. Proceedings of the 42nd
International ACM SIGIR Conference on Research
and Development in Information Retrieval, 235–244.
11. Dou, Y., Yu, P. S., Ren, X., Sun, Y., & Yu, R.
(2019). Enhancing recommender systems with
knowledge-aware attention network. Proceedings
of the 2019 World Wide Web Conference (WWW),
2009–2020.
12. Sun, X., Zhang, W., & Zhang, X. (2020). Multimodal
sentiment-enhanced recommendation.
Proceedings of the 43rd International ACM SIGIR
