Figure 1 - uploaded by Paul Marrow
Content may be subject to copyright.
Source publication
Technical report on field trial of recommender system used for marketing video-on-demand distributed over network by BT Group plc. Part of output from collaborative project MyMedia. Success of this field trial led to product development now supporting personalised recommendation for a number of online TV services.
Contexts in source publication
Context 1
... this context the presentation of editorial recommendations does have an advantage, which may explain the substantial difference between the groups in August, where the hit-rate recorded for the Editorial group is higher. August is a month for holidays, and as Table 4 and Figure 10 above show, is a time for higher levels of tVoD viewing on the BT Vision service. Accordingly it is a time for considerable marketing effort, and the recommendations provided by the editorial team for the BT MyMedia field trial Editorial recommendations are consistent with those marketed by other means. So even if members of the Editorial trial group have not viewed the recommendations page, and the evidence from section 8.3.3 suggests that that is the case for most members of the group, they may have seen similar items marketed by other means. By contrast the MyMedia recommender algorithm does not drive any additional marketing ...
Context 2
... reported in deliverable D1.5 of the MyMedia project [1], and summarised in a paper by Knijnenburg et al. (2010) [2], led to the development of a conceptual framework for evaluating recommender systems intended to be sufficiently general as to be able to cover all the field trials in the MyMedia project, as well as other recommender systems as well. For the sake of rapid comparison, the a simplified version of the framework as shown in Figure 10 of deliverable D1.5 is reproduced in Figure 9 ...
Similar publications
This paper proposes the implicit social networking as an innovative methodology for approaching consumers who possess information-rich user profiles based on aplethora of online services they use. An implicit social network is not explicitly built by consumers themselves, but implicitly calculated by third parties based on a level of a common inter...
Citations
... MyMediaLite is based on parts of the framework [16] that has been used in four different industrial field trials of the MyMedia project, including one involving 50,000 households [17]. Application areas in the field trials were IPTV, webbased video and audio, and online shopping. ...
MyMediaLite is a fast and scalable, multi-purpose library of recommender system algorithms, aimed both at recommender system researchers and practitioners. It addresses two common scenarios in collaborative filtering: rating prediction (e.g. on a scale of 1 to 5 stars) and item prediction from positive-only implicit feedback (e.g. from clicks or purchase actions). The library offers state-of-the-art algorithms for those two tasks. Programs that expose most of the library's functionality, plus a GUI demo, are included in the package. Efficient data structures and a common API are used by the implemented algorithms, and may be used to implement further algorithms. The API also contains methods for real-time updates and loading/storing of already trained recommender models.
MyMediaLite is free/open source software, distributed under the terms of the GNU General Public License (GPL). Its methods have been used in four different industrial field trials of the MyMedia project, including one trial involving over 50,000 households.
