RecDelta

An Interactive Dashboard on Top-K Recommendation for Cross-model Evaluation

Datasets
Algorithms
Item Metadata
Example: { "attr": { "name": "title", "nominal": ["genres"], "numerical": ["release_time"] }, "data": { "item_1": {"title": "Toy Story (1995)", "release_time", "1995", "genres": ["Adventure", "Animation", "Children"], "item_2": {"title": "3 Idiots"}, "genres": ["Comedy", "Fantasy"], .... } }
User History
Example: {"user_1": ["item_2492", "item_2012", "item_2478", "item_553", "item_1445", ...], "user_2": ["item_2943","item_3053", "item_3273", ...], ...}
(Items should be sorted by interaction time point.)
Embeddings for Model [[index+1]]
Example:
item_1 \t 1.123 -120.0 144 \n
item_2 \t 0.123 -10.8 57.2 \n
...
Example:
user_1 \t 1.123 -120.0 144 \n
user_2 \t 0.123 -10.8 57.2 \n
...