Project Paper: https://www.overleaf.com/read/hbzdkxfpjhkx
A movie recommender system based on Information Retrieval and Data Integration Theory. Uses content-based matching to link a user's Twitter profile contents to movie information stored in a local MongoDB instance. These results are scored using a TF-IDF style weighting system, and displayed within a GUI designed using Python's tkinter library.
Three modes of recommendation:
- Twitter: Recommendations based on a user's Twitter profile contents.
- History: Recommendations based on a list of movies specified by the user (such as a watch history)
- Tags: Recommendations based on a list of tags specified by the user (genres, topics, etc)
- MovieLens Latest Datasets
- MovieLens Tag Genome Dataset
- Facebook API
- Offline IMDB Data
- Online IMDB API
- Tweepy: Twitter for Python!
- Twitter REST APIs
- Affective Norms for English Words
- TextBlob: Simplified Text Processing
- AYLIEN Text Analysis API
- omdb.py: Python wrapper around The Open Movie Database API
