Table 8 Application focused and dataset description in different recommender systems
From: A systematic review and research perspective on recommender systems
References | Application | Dataset description |
|---|---|---|
Castellano et al. [1] | Usage-based Web recommender system | Datasets consist of real-world data gathered from the log-files of a commonly visited website of the successful Japanese film: Dragon Ball Dataset size: 434 input–output samples |
Crespo et al. [2] | Intelligent electronic books | – |
Lin et al. [3] | Automatic vending machines | Data from nine vending machines were collected from two locations |
Wang and Wu [4] | Ubiquitous-learning system | Botanical database Dataset size: 80 knowledge units |
García-Crespo et al. [5] | Hotels | Data collected from 10 hotels in Mallorca (Spain) |
Dong et al. [6] | Business and software environment | Service metadata Dataset size: 1000 |
Li et al. [7] | Knowledge management systems for the aircraft industry | – |
Lorenzi et al. [8] | Tourism domain | A private dataset was collected from a travel agency Dataset size: 35 test cases |
Huang et al. [9] | Context-aware recommender system for movie recommendation | Set of real-world movie rating data Dataset size:1048 ratings from 116 users |
Chen et al. [10] | Anti-diabetic drugs selection | Database collected from a professional community of America based physicians Dataset size: 20 patients |
Mohanraj et al. [11] | Online recommender system | Data sourced from an intranet website |
Hsu et al. [12] | Recommendations for auxiliary material | Facebook data Dataset size: 10 datasets each having 200 to 2000 Facebook posts |
Gemmell et al. [13] | Resource recommender for social annotation systems | Data sourced from: Bibsonomy (size: 13,909 annotations), Citeulike (size: 105,873 annotations), movielens (size: 35,366 annotations), Delicious (size: 720,788 annotations), Amazon (size: 498,217 annotations) and lastfm (size: 172,177 annotations) |
Choi et al. [14] | Online-product recommender system | Data was collected from a major online shopping mall in Korea Dataset size: 16,486 transactions |
Garibaldi et al. [15] | Breast cancer treatment recommendation | Data was collected from a hospital in the United Kingdom Dataset size: 1310 patients |
Salehi and Kmalabadi [16] | E-learning material recommender system | Open-source data gathered from a course management system Dataset size: 16,345 records |
Aher and Lobo [17] | E-learning recommender system | Open-source data gathered from a course management system |
Kardan and Ebrahimi [18] | A content recommender system for asynchronous discussion groups | MetaFilter dataset (size: 229,401 threads) and Yahoo Answers dataset (size: 7,941,404 threads) |
Chang et al. [19] | A cloud-based intelligent TV program | User profile database Dataset size: 300,000 user profiles |
Lucas et al. [20] | Tourism recommender system | Data was collected from a personalized tourist-aiding system Dataset size: 241 |
Niu et al. [21] | Video recommender system | Metadata of videos Dataset size: 362 videos |
Liu et al. [22] | Recommender system for self-driven tourists | Data collected from Google Maps |
Bakshi et al. [23] | E-commerce recommender system | MovieLens dataset of 100,000 ratings provided by 943 users on 1682 movies |
Kim and Shim [24] | Data fetched from Twitter Dataset size: 12,098,339 tweet messages | |
Wang et al. [25] | Movie recommender system | MovieLens dataset of 100,000 ratings by 943 users on 1682 movies |
Kolomvatsos et al. [26] | Music and book recommender system | An open-source song dataset Dataset size: 1167 songs |
Gottschlich et al. [27] | Stock investment recommender system | Data were collected from the 30 largest public companies in Germany |
Torshizi et al. [28] | Benign prostatic hyperplasia | 44 patients’ data were collected from a Medical University |
Zahálka et al. [29] | Venue recommender | New York dataset (size: 7,246 venues and 1,072,181 items), and Amsterdam dataset (size: 693 venues and 55,990 items) |
Sankar et al. [30] | Stock recommender system | Data were collected from 17 mutual funds |
Chen et al. [31] | Movie recommender system | Data was collected from the MovieLens dataset of 100,000 ratings collected from 943 users on 1682 movies and eachmovie dataset with 2,811,983 ratings from 61,265 users on 1623 movies |
Wu et al. [32] | Item recommender system | Delicious (Dataset size: 69,223 items), Lastfm (Dataset size: 12,523 items) and CiteULike (Dataset size: 7,188 items) |
Yeh and Cheng [33] | Tourist management | Tourist related data was collected from 457 users |
Liao et al. [34] | e-commerce recommender system | Data was collected from many online shopping platforms |
Li et al. [35] | Movie recommender system | Netflix Prize dataset (size: 100,480,507 ratings from 480,189 customers on 17,770 movies) |
Wu et al. [36] | Social media recommender system | Delicious (Dataset size: 45,419 items) and Lastfm (Dataset size: 12,523 items) |
Adeniyi et al. [37] | Web usage-based recommender system | Web server data was collected from 5285 web pages |
Rawat and Kankanhalli [38] | Mobile photography | Flickr API Dataset size: 67,000 images |
Yang et al. [39] | Job recommender system | 4 months’ data were extracted from the job posting and job application records |
Lee et al. [40] | Music streaming recommender system | An open-source song dataset |
Wei et al. [41] | Cold start items recommender system | Large Netflix dataset with 100 million ratings for 17,770 movies given by 480,189 users |
Li et al. [42] | Diverse applications | MovieLens dataset of 1 million ratings on 3952 movies from 6040 users |
Mezei and Nikou [43] | Health and wellness recommender systems | Publicly available medical dataset Dataset size: 20,560 data points |
Ayata et al. [44] | Music recommender system | Multimodal DEAP emotion database Dataset size: 32 items |
Zhao et al. [45] | Social-aware movie recommender system | A Chinese movie recommending website has 99,641 ratings on 59,695 movies collected from 4,242 users |
Hammou et al. [46] | Big Data recommender system | Two MovieLens datasets of size 10 million and 20 million ratings respectively, and Yelp dataset with 6,685,900 ratings |
Zhao et al. [47] | Social network recommender system | Two MovieLens datasets of size 100,000 and 1 million ratings respectively |
Bhaskaran and Santhi [48] | E-learning and cloud computing recommender system | A public Book-Crossing dataset with 2,78,858 users providing 1,149,780 ratings on 271,379 books |
Afolabi and Toivanen [59] | Chronic disease management recommendation | Health-related data collected from devices and sensors |
He et al. [60] | Movie recommender system | MovieLens dataset Dataset size: 1,000,000 movie ratings |
Han et al. [49] | Cancer rehabilitation recommender system | Cancer rehabilitation medical data |
Kang et al. [50] | Advertisement recommender system for online broadcasting | Amazon product data has data from 1557 users, 79,344 reviews, and 55,101 items |
Ullah et al. [51] | Image-based service recommendation | Amazon product image data Dataset size: 3.5 million items |
Cai et al. [52] | General-purpose recommender system | MovieLens dataset Dataset size: 1,000,000 movie ratings |
Esteban et al. [53] | University course selection recommendation | Student data was collected from a university Dataset size: 2500 items |
Mondal et al. [54] | Doctor recommender system | Neo4j NoSQL graph database, and patient-doctor information collected from different hospitals Dataset size: 5300 patient records |
Dhelim et al. [55] | Product recommender for online shopping | Integrated dataset of a product recommendation system and a social network platform Dataset size: 6230 items |
Bhalse et al. [56] | Movie recommender system | MovieLens dataset containing 100,837 ratings applied over 9743 movies |
Ke et al. [57] | Cross-platform dynamic goods recommender system | Data was collected using a cross-platform data collection API Dataset size: 100 users |
Chen et al. [58] | Video recommender system | MovieLens dataset of size 10 million items and micro-video dataset of size 1.7 million items |