click here to download the project abstract of movie recommendation
At DataPro, we provide final year projects with source code in python for computer science students in Hyderabad , Visakhapatnam.
ABSTRACT
A movie recommendation is important in our social life because it has the ability to increase the entertainment for the people. every year there are many movies released and the movie lovers show different choices to particular movies. From the pool of movies user finds it difficult to select and watch the movies of his or her interest.so there comes the need of movie recommendation engine. Recommendation systems made finding things easy that the users need. Most of the existing movie recommendation systems generally provide the overall rating of the movie without any personalization
of users.
we have implemented and evaluated content based and collaborative basedm filtering on movie lens dataset by analysing both the models we tried to build a hybrid model to increase the accuracy of the movies recommended to the user. This engine aims for personalization of a user and ease the process of searching of movies based on users interests there by reducing human efforts. Nowadays the recommendation system has made finding the things easy that we need. Most existing movie services like IMDB do not personalize their recommendations but simply provide an overall rating for a movie. This significantly decreases the value of each recommendation as it does not cater to the individual movie preferences of the user. Unlike these systems, our proposed Recommendation Engine will continually
analyse individual user’s movie preferences and recommend custom movie recommendations. This aims for personalization of a user, and the overall goal is to ease the movie discovery process. It reduces human effort by suggesting movies based on the user’s interests
INTRODUCTION
Basically, there exists some dependency between user and movies they like to watch. An efficient recommendation engine should explore these both dependencies and recommend related movies to the users. Many companies are spending billions ofdollars on implem enting an accurate and personalized recommendation algorithm. Best example for this is Netflix prize competition. This was an open competition for the best collaborative filtering algorithm to find the ratings of unrated movies of the user basing on his previously rated movies without any other data. The competition was held by Netflix in 2006.mostly recommendation systems are of two types which are based on content based both filtering and collaborative filtering.
it is like if a user likes a movie A, he will also like similar movies like A. Collaborative filtering filters information by using the interactions and data collected by the system from other users. It’s based on the idea that people who agreed in their evaluation of certain items are likely to agree again in the future. The evolution of technology brings us many advanced platforms such as Machine Learning, Deep Learning, Internet of Things, etc. We are using these technologies to satisfy our needs. Recommendation system is the best example for this. Recommendation system is using Machine Learning to recommend Movies, Songs, Products in E-Commerce websites, etc. Methods which are widely used in Recommendation System are Content based filtering, Collaborative filtering. Collaborative filtering is one of the most popular methods to implement a recommender system. Collaborative filtering is of two types
Memory based approach and Model based approach.
Dataset:
We have used movie lens 100k dataset. The data set consists of
- 100,000 ratings (1-5) from 943 users on 1682 movies.
- Each user has rated at least 20 movies.
The data was collected through the movie lens website
K NEAREST NEIGHBOUR ALGORITHM(KNN):
Step 1-Collecting the data:
importing the data set (users, movie id, ratings) from any sources like movie lens or Kaggle. Since we want to implement using python, Import the required Python
libraries like Pandas, NumPy, Seaborn, and Matplotlib.
Step 2-Data cleaning or filtering:
In a real-world setting, data collected from explicit feedback like movie ratings can be very sparse and data points are mostly collected from very popular items (movies) and highly engaged users. Large amounts of less known items (movies) don’t have ratings at all.
Step 3-Finding similarity between different users:
- Select a user and the movies that the user watched
- Find users who have also watched some or all of those movies which are
watched by the selected user. - For each user, extract ratings for similar movies.
- Calculate similarity score using cosine similarity or Pearson correlation which
is invariant to scaling. - Find top X users.
- Find weighted rating of top X users to all of their watched movies.
- Finally add up correlation, weighted ratings for each movie watched by more
than top X users. - For each movie, find (weighted ratings/correlation).
- Sort these values.
- Finally extract top movies to recommend
In user based the algorithm finds X most similar users to user A and uses them for basis of recommendation. Item based collaborative filtering was introduced 1998 by Amazon. Unlike user based collaborative filtering, item-based filtering looks at the similarity between different items, and does this by taking note of how many users that bought item X also bought item Y.
