Technical Areas:
- Used pandas and NumPy to organize the data in an ordered set. Then used pandas dataframe functions to perform data manipulation and clean the data.
- Gained additional insight into vectorization and cosine similarity
- Understood the mechanism of the recommender system
- Implemented a movie recommender system from scratch based on these learnings
Tools:
- Jupyter Notebook
- Spacy, Pandas, NumPy, CSV, NLTK
- Classifiers used: K-nearest neighbours, Cosine Similarity
- Git / GitHub
Soft Skills:
- Gained insight into ML concepts and functions using tutorials and online videos
- Explored pandas data frame functions and debugged code
- Utilized new libraries and read their documentation throughly
3 Achievement Highlights:
- Created basic recommender system with cosine similarity
- Understood the logic behind recommender systems and ML algorithms
Tasks Completed:
- Removed stopwords and punctuations from the text data
- Implemented a basic movie recommender system using cosine similarity (as recommended)
- Categorized the movies based on genres and descriptions to determine 10 most similar movies
Link to project: https://github.com/harshita19244/classifiers/blob/main/recommender.ipynb