Collaborate Filtering for Digital Publishing |
Rasmus Theodorsen
|
| Abstract | Recommender systems provide users with personalized suggestions for products or services. The recommendations can be based on the nature of the products or it can be based on collaborate filtering. Collaborate filtering takes the behaviour and patterns of users into account in order to make recommendations that reflects the preferences of the users, inferred by their past behaviour.
The goal of the thesis is to test collaborate filtering models which could be used as part of a recommender system for the publication portal, Issuu. To do so two models have been tested on a data set provided by Issuu. The data set
contains the publications read by a user as well as the amount of time the user has spent on the publication. By applying the read time as an implicit rating of a publication the models were tested for their ability to predict how a user would rate an item.
Finally it is shown how each of the models can be adapted into a live recommender system, where the models are to be able to output recommendations for a given user. It is investigated how the models can handle new data entries as they are generated on the website, and how the models should treat entries as they age. |
| Type | Master's thesis [Industrial collaboration] |
| Year | 2013 |
| Publisher | Technical University of Denmark, Department of Applied Mathematics and Computer Science / DTU Co |
| Address | Matematiktorvet, Building 303B, DK-2800 Kgs. Lyngby, Denmark, compute@compute.dtu.dk |
| Series | M.Sc.-2013-60 |
| Electronic version(s) | [pdf] |
| Publication link | http://www.compute.dtu.dk/English.aspx |
| BibTeX data | [bibtex] |
| IMM Group(s) | Image Analysis & Computer Graphics |