Description:
We are encountering an issue where the same story is recommended multiple times in our clustering-based recommendation engine. This problem occurs in the recommend_stories function where stories from a user's preferred cluster and the most popular cluster are combined, leading to potential duplicates in the final recommendation list.
Tasks:
[x] Previously, top_recommendations and top_cluster_stories were concatenated and the duplicates were dropped. drop_duplicates was removed from combined_recommendations function since it wasn't working as expected.
[x] Instead implement drop_duplicates on the recommended_stories DataFrame to make sure each recommended story is unique.
JohnsL-U
commented
11 months ago
Description: We are encountering an issue where the same story is recommended multiple times in our clustering-based recommendation engine. This problem occurs in the recommend_stories function where stories from a user's preferred cluster and the most popular cluster are combined, leading to potential duplicates in the final recommendation list.
Tasks:
[x] Previously, top_recommendations and top_cluster_stories were concatenated and the duplicates were dropped. drop_duplicates was removed from combined_recommendations function since it wasn't working as expected.
[x] Instead implement drop_duplicates on the recommended_stories DataFrame to make sure each recommended story is unique.