forked from BloomTech-Labs/Groa-ds
-
Notifications
You must be signed in to change notification settings - Fork 1
/
notes
21 lines (18 loc) · 1.42 KB
/
notes
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
IMDB:give imdb username, search database for reviews with that username. Then do the modeling on those reviews.
Letterboxd: upload zip file, unzip, retrieve reviews and ratings files
test list: {"list":"3140100,3147312,5534434,0095215,8504014,8535180,0095583,1591095,0095925"}
Things we need from front end:
- Username (IMDb and Letterboxd(hopefully)) -- Page: Import Reviews (maybe homepage)
- Manually put in a review --- Page:Submit Reviews (needs movie title, review, (review title) title predictor dropdown(film not here)?
- user input: username, user_rating, review_title
- dropdown: original_title or primary_title
- View recommendations from model - persist on the recommendations we have served to each user ((issue for future web team))
- have a page that returns recommendations
- Whether they like or dislike recommendations (thumbs up or down) (either add to watchlist or won't watch)
- View watch history
- Table? Include "results for username watch history" as title, columns: primary_title, review_date, review_title (clickable link to full review text)
- make it clickable so you can mix and match which movies they want to server to the model
- feature it as thumbnails
- Serve users their previous reviews (show them what they've already reviewed from other places.)
- import ratings and use 7-10 as stuff for preliminary recommendation
-then give list and give options for serving to recommender