Output details
15 - General Engineering
Brunel University London
Semantic content ranking through collaborative and content clustering
This paper reports on our early attempts at exploiting evolutionary computation techniques in semantic content modelling and, in particular, in enhancing our EPSRC-funded semantic content modelling system, COSMOS-7. Until then, when queried, COSMOS-7’s output was a sequence of relevant yet unranked video segments which users have had to sift through. Using Self-Organising Neural Networks, we developed an add-on module that clusters and ranks COSMOS-7 output through consideration of user preferences and knowledge gained from usage of the same content by similar users and similar content by the same user. The research was funded through EPSRC doctoral training accounts (EP/P501334/1, GR/P01496/01).