GithubHelp home page GithubHelp logo

recsys_data_zoo's Introduction

Recommender System Data Collection

List of Dataset


Dataset Domain Feedback Timestamp Auxillary information Link
ML-1M Movie Rating O User, movie metadata (e.g. age, gender, genre, …) Page
ML-10M Movie Rating O User, movie metadata (e.g. age, gender, genre, …) Page
ML-20M Movie Rating O User, movie metadata (e.g. age, gender, genre, …) Page
ML-25M Movie Rating O User, movie metadata (e.g. age, gender, genre, …) Page
Amazon Music (5-core) E-commerce Rating O Review text, user, item metadata (e.g. useful votes, style) Page
Amazon CD (5-core) E-commerce Rating O Review text, user, item metadata (e.g. useful votes, style) Page
Amazon Games (5-core) E-commerce Rating O Review text, user, item metadata (e.g. useful votes, style) Page
Amazon C&A (5-core) E-commerce Rating O Review text, user, item metadata (e.g. useful votes, style) Page
Ciao (5-core) E-commerce Click O Review text, user, item metadata (e.g. useful votes, style) Page
Epinions E-commerce Rating X - Download
Yelp-2015 Restaurant Rating - - (Original page is closed) Page
Yelp-2018 Restaurant Rating - - (Original page is closed) Page
CiteULike-a Citation Network Click X Tags of item Page
Pinterest Social Network Click X Item metadata (e.g. Item category) Raw, NCF
Gowalla Social Network Click O Relation between users, item metadata (e.g. longitude, latitude) Page

Preprocess Method


K-minimum User Interactions

- Filter out users with less than K interactions
- Filter out items with no interactions

K-core Setting

- Filter out users with less than K interactions
- Filter out items with less than K interactions
- Repeat until there is no further change in # users and items

Statistics


Definitions

- User Avg. Ratings: (# Ratings) / (# Users)
- Sparsity: 1 - {(# Ratings) / (# Users * # Items)}
- Shape: # Users / # Items
- Gini User/Item: Gini-index of user/item popularity. Ratings are more inequally distributed as the index is close to 1.
- Item Concen.: (Item Concentration) Ratio of the ratings that top 5% items hold.

Raw data (data without preprocessing)

Dataset # Users # Items # Ratings User Avg. Ratings Sparsity Shape Gini User Gini Item Item Concen.
ML-1M 6,040 3,706 1,000,209 165.60 0.9553 1.6298 0.5286 0.6336 0.2828
ML-10M 69,878 10,677 10,000,054 143.11 0.9866 6.5447 0.5707 0.8052 0.5165
ML-20M 138,493 26,744 20,000,263 144.41 0.9946 5.1785 0.5807 0.9029 0.7141
ML-25M 162,541 59,047 25,000,095 153.81 0.9974 2.7527 0.5895 0.9419 0.8445
Amazon Music 16,566 11,797 145,292 8.77 0.9993 1.4043 0.3756 0.4315 0.2476
Amazon CD 112,395 73,713 1,402,148 12.48 0.9998 1.5248 0.4656 0.5432 0.3353
Amazon C&A 157,212 48,186 1,120,771 7.13 0.9999 3.2626 0.2091 0.5875 0.3787
Amazon Games 55,223 17,408 473,427 8.57 0.9995 3.1723 0.3230 0.5849 0.3484
Epinions 40,163 139,738 664,823 16.55 0.9999 0.2874 0.6763 0.6936 0.5339
Ciao 6,762 16,610 146,997 21.74 0.9987 0.4071 0.5518 0.5309 0.3128
Yelp 2015 25,677 25,815 696,865 27.14 0.9989 0.9947 0.4509 0.6037 0.3512
Yelp 2018 45,919 45,538 1,183,609 25.78 0.9994 1.0084 0.4268 0.5810 0.3458
Gowalla 107,092 1,280,969 3,981,334 37.18 0.9999 0.0836 0.6627 0.5390 0.3628
CiteULike 5,551 16,980 204,986 36.93 0.9978 0.3269 0.4706 0.3696 0.2098
Pinterest 55,187 9,916 1,463,580 26.52 0.9973 5.5655 0.1401 0.4511 0.1899

10-minimum User Interactions

Dataset # Users # Items # Ratings User Avg. Ratings Sparsity Shape Gini User Gini Item Item Concen.
ML-1M 6,040 3,706 1,000,209 165.60 0.9553 1.6298 0.5286 0.6336 0.2828
ML-10M 69,878 10,677 10,000,054 143.11 0.9866 6.5447 0.5707 0.8052 0.5165
ML-20M 138,493 26,744 20,000,263 144.41 0.9946 5.1785 0.5807 0.9029 0.7141
ML-25M 162,541 59,047 25,000,095 153.81 0.9974 2.7527 0.5895 0.9419 0.8445
Amazon Music 3,951 11,483 75,044 18.99 0.9983 0.3441 0.3067 0.4505 0.2405
Amazon CD 36,487 73,493 933,651 25.59 0.9997 0.4965 0.4448 0.5455 0.3296
Amazon C&A 21,361 44,416 301,976 14.14 0.9997 0.4809 0.1915 0.5807 0.3479
Amazon Games 11,986 17,100 215,645 17.99 0.9989 0.7009 0.2994 0.5833 0.3319
Epinions 15,786 132,964 580,752 36.79 0.9997 0.1187 0.4816 0.6768 0.5197
Ciao 3,829 15,768 129,798 33.90 0.9979 0.2428 0.4516 0.5267 0.3119
Yelp 2015 24,930 25,799 690,381 27.69 0.9989 0.9663 0.4482 0.6038 0.3510
Yelp 2018 45,842 45,538 1,182,917 25.80 0.9994 1.0067 0.4267 0.5811 0.3458
Gowalla 68,709 1,247,158 3,831,386 55.76 0.9999 0.0551 0.5458 0.5365 0.3612
CiteULike 5,551 16,980 204,986 36.93 0.9978 0.3269 0.4706 0.3696 0.2098
Pinterest 55,187 9,916 1,463,580 26.52 0.9973 5.5655 0.1401 0.4511 0.1899

20-minimum User Interactions

Dataset # Users # Items # Ratings User Avg. Ratings Sparsity Shape Gini User Gini Item Item Concen.
ML-1M 6,040 3,706 1,000,209 165.60 0.9553 1.6298 0.5286 0.6336 0.2828
ML-10M 69,878 10,677 10,000,054 143.11 0.9866 6.5447 0.5707 0.8052 0.5165
ML-20M 138,493 26,744 20,000,263 144.41 0.9946 5.1785 0.5807 0.9029 0.7141
ML-25M 162,541 59,047 25,000,095 153.81 0.9974 2.7527 0.5895 0.9419 0.8445
Amazon Music 1,074 10,116 37,526 34.94 0.9965 0.1062 0.2673 0.4466 0.2298
Amazon CD 12,284 71,838 616,845 50.22 0.9993 0.1710 0.4262 0.5472 0.3234
Amazon C&A 2,302 25,285 66,772 29.01 0.9989 0.0910 0.1976 0.4593 0.2585
Amazon Games 2,734 15,189 97,227 35.56 0.9977 0.1800 0.2826 0.5611 0.3069
Epinions 8,693 123,330 482,849 55.54 0.9996 0.0705 0.4192 0.6525 0.4983
Ciao 2,075 14,887 105,827 51.00 0.9966 0.1394 0.3814 0.5224 0.3083
Yelp 2015 9,788 25,373 489,820 50.04 0.9980 0.3858 0.3972 0.5959 0.3346
Yelp 2018 17,137 45,447 806,078 47.04 0.9990 0.3771 0.3768 0.5853 0.3381
Gowalla 47,752 1,183,848 3,530,010 73.92 0.9999 0.0403 0.4946 0.5289 0.3558
CiteULike 3,097 16,792 171,391 55.34 0.9967 0.1844 0.3939 0.3873 0.2139
Pinterest 52,190 9,909 1,408,089 26.98 0.9973 5.2669 0.1359 0.4578 0.1923

10-core Settings

Dataset # Users # Items # Ratings User Avg. Ratings Sparsity Shape Gini User Gini Item Item Concen.
ML-1M 6,040 3,260 998,539 165.32 0.9493 1.8528 0.5285 0.5862 0.2594
ML-10M 69,878 9,708 9,995,471 143.04 0.9853 7.1980 0.5706 0.7867 0.4945
ML-20M 138,493 15,451 19,964,833 144.16 0.9907 8.9634 0.5802 0.8360 0.5784
ML-25M 162,539 24,330 24,890,566 153.14 0.9937 6.6806 0.5881 0.8711 0.6558
Amazon CD 21,450 18,398 527,503 24.59 0.9987 1.1659 0.4286 0.4274 0.2351
Amazon Games 5,942 3,793 103,778 17.47 0.9954 1.5666 0.2857 0.4028 0.2158
Epinions 10,706 8,945 300,303 28.05 0.9969 1.1969 0.4037 0.4950 0.3055
Ciao 2,136 2,597 59,884 28.04 0.9892 0.8225 0.3982 0.4063 0.2855
Yelp 2015 22,087 14,873 602,517 27.28 0.9982 1.4850 0.4413 0.5121 0.2935
Yelp 2018 39,055 25,033 988,768 25.32 0.9990 1.5601 0.4168 0.5065 0.2888
Gowalla 29,858 40,988 1,027,464 34.41 0.9992 0.7285 0.4666 0.4346 0.2915
CiteULike 3,710 6,468 120,324 32.43 0.9950 0.5736 0.4388 0.3052 0.1818
Pinterest 55,164 9,316 1,460,487 26.48 0.9972 5.9214 0.1411 0.4188 0.1819
  • All users and items are deleted in Amazon-{music, C&A}

recsys_data_zoo's People

Contributors

yoongi0428 avatar

Stargazers

 avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar  avatar

Watchers

 avatar

Recommend Projects

  • React photo React

    A declarative, efficient, and flexible JavaScript library for building user interfaces.

  • Vue.js photo Vue.js

    🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.

  • Typescript photo Typescript

    TypeScript is a superset of JavaScript that compiles to clean JavaScript output.

  • TensorFlow photo TensorFlow

    An Open Source Machine Learning Framework for Everyone

  • Django photo Django

    The Web framework for perfectionists with deadlines.

  • D3 photo D3

    Bring data to life with SVG, Canvas and HTML. 📊📈🎉

Recommend Topics

  • javascript

    JavaScript (JS) is a lightweight interpreted programming language with first-class functions.

  • web

    Some thing interesting about web. New door for the world.

  • server

    A server is a program made to process requests and deliver data to clients.

  • Machine learning

    Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.

  • Game

    Some thing interesting about game, make everyone happy.

Recommend Org

  • Facebook photo Facebook

    We are working to build community through open source technology. NB: members must have two-factor auth.

  • Microsoft photo Microsoft

    Open source projects and samples from Microsoft.

  • Google photo Google

    Google ❤️ Open Source for everyone.

  • D3 photo D3

    Data-Driven Documents codes.