Abstract
Reviews are essential for any company’s improvement and growth, but some of them may not be very useful and as the number increases they can be very hard to manage. For this project, we work with Ocado research team to develop a Machine Learning model that is capable of applying Natural Language Processing (NLP) techniques, analysing the review text and predicting the rating and helpfulness of this particular review. It can help to clean the unhelpful reviews, correct the reviews that are tagged with wrong ratings, and will produce some useful methodology and learning patterns for the research team.
Key Features
- Supervised learning model that improves by feeding in example data with correct answers.
- Use NLP to process reviews.
- Can predict the rating and helpfulness based on the text of the review.
- Written in Python.
Development Team
Krzysztof Kozinski (Team leader, programmer)
zcabkoz@ucl.ac.uk
Walter Wu (Main programmer, website editor)
zcabwux@ucl.ac.uk
Rui Chen (Programmer, poster editor)