2012-07-09: first published.
2015-09-23: published on www.lucami.org
The LDOS-CoMoDa is a context rich movie recommender dataset. It contains ratings for the movies and the twelve pieces of contextual information describing the situation in which the movies were consumed. Some of the important dataset properties are: the ratings and the contextual information are explicitly acquired from the users immediately after the user consumed the item; contextual information describes the situation in which the user consumed the item; the ratings and the contextual information are from real user-item interaction and not from hypothetical situation or user’s memory of past interactions; users are able to rate the same item more than once if they consume the item multiple times.
The dataset is acquired for the needs of our research on context-aware recommender systems. We are glad to share this dataset with the other researchers. We believe that the other members of the research community will find the data valuable for their research.
The data is still being acquired and the new versions will be available.
A scinetific paper providing further detais: ODIĆ, Ante, TKALČIČ, Marko, TASIČ, Jurij F., KOŠIR, Andrej: Predicting and Detecting the Relevant Contextual Information in a Movie-Recommender System accessible here.
The dataset is presented in the paper: KOŠIR, Andrej, ODIĆ, Ante, KUNAVER, Matevž, TKALČIČ, Marko, TASIČ, Jurij F. Database for contextual personalization. Elektrotehniški vestnik. [English print ed.], 2011, vol. 78, no. 5, str. 270-274, ilustr. [COBISS.SI-ID 8871764]
The description of the data in the dataset can be found at LDOS-CoMoDa-description .
To access the LDOS-CoMoDa dataset please contact: email@example.com
Extract form the dataset