This project optimizes the review highlight section in Yelp by tailoring suggestions according to user taste-profile as well as review informativeness. By applying collaborative filtering techniques, it uncovers hidden tastes of Yelp reviewers using their review history as well as restaurant attributes. A cross-validated predictive model automatically selects the three reviews with the highest likelihood of being tagged as "useful" by the community, and presents them in descending order according to user-similarity.
alexruize/YelpReviewRank
Folders and files
| Name | Name | Last commit date | ||
|---|---|---|---|---|