Recommender Systems

Pavithra Brahmananda Reddy

Name: Pavithra Brahmananda Reddy
Majors: Computer Science, Mathematics
Advisor: Thomas Montelione
Second Reader: Robert Kelvey

Many products that we buy online are suggested to us using the power of recommendation systems. We might be able to observe this with products on Amazon and content on Netflix. Using efficient recommendations has been proven to improve business and make iteasier for customers to choose from the seemingly never-ending list of choices. The goal of my Independent Study project was to explore the different approaches to make recommendations to users. The focus of this project was to provide relevant movie recommendations. The paper provides an introduction to recommendation systems and their increased relevance in the recent past. It also discusses the history behind the development of recommendation systems and how they have improved over time. Further, it discusses the most commonly used approaches along with their strengths and weaknesses. Data mining and natural language processing techniques that are used to generate recommendations have been examined in detail. These algorithms have been implemented in theweb-based application. Collaborative and content-based filtering are the two main types of recommendation systems that were studied. This research provides insight into the cutting-edge technology being used by streaming services to attract and retain customers. In addition to this, working on this project provided the opportunity to gain experience in application development. The performance of the recommendation system was evaluated by using a part of the dataset to test the predictions and measure accuracy. Further, a survey was conducted to obtain feedback about user experience with the application.

Posted in Comments Enabled, Independent Study, Symposium 2022.

12 responses to “Recommender Systems”

  1. Maris says:

    This system is amazing my top recommendation was Beetlejuice!

  2. Brahma says:

    Today Artificial Intelligence is touching all facets of our life. Recommender Systems play a crucial role in e-commerce websites, helping us find easily movies we may like to watch, products that may be of interest to us..As I understand from the write-up NLP and SVD are the two main technologies underpinning such text based recommendation engines. I suppose this must have been both challenging and highly fulling experience. Hearty Congratulations for this excellent work..

  3. Pramod Samal says:

    Very nicely explained . Hearty Congratulations for the deep understanding and more importantly explaining the concept in Simple terms . Truly amazing work .

  4. Vivekanand Vadodkar says:

    Congratulations. This topic is very relevant in the current context. Handled very well by the authors.

  5. Vimala says:

    It is fantastic.. Hearty Congratulations..

  6. Pankaj Bharadia says:

    May I suggest to use A B analysis. Always provide choices, say 2 versions A and B to users and based on their feed backs, continue to improve the system.

    Job of AI/ML is to continue to learn and improve based on analysis of logs as well as feedback.

    Even some irrational random recommendations should be inserted into the flow to see how users react and keep on learning.

    Congratulations, well done Pavithra.

  7. Manoj Tarade says:

    Good work Pavithra, personally as a user i see benefits in this AI based recommendations, it’s useful. One tricky requirement could be: Once buying is done then such recommendations should go away as well.

    Keep up your good work and take it next ok level. Do well and all the best!

  8. Ram Dubakula says:

    Very great achievements. Nicely written project. Congratulations, well done Pavithra

  9. Rajesh Bhalerao says:

    Nice work and congratulations!

    It would be helpful to understand if any open-source models were used to build this system.

    All the best.

  10. Jai Singh Baghel says:

    Great job Pavithra…many congratulations

  11. Maha Rashid says:

    Super interesting IS!

  12. Max Johnson (Morbius Ticket Owner) says:

    Congrats on your project! I hope that one day this can be adapted to recommend only Shrek and Shrek-adjacent movies to properly culture our future generations, it’s good to enlighten our fellow humans. Very Larva