Personalisation, Recommender Systems
SSE - CREST Workshop October 1, 2010
1. introductions
●   licia capra

●   neal lathia (me)
●   giovanni quattrone

●   michalis christodoulou
●   afra mashhadi
●   lucia del prete
●   valentina zanardi
●   more tba
2. research perspective
people
i seek   places
         things
         events
         information
connect   information
          systems
          people,places,
          things,events,info
information
               systems
extract from   people,places,
               things,events,info
information
            systems
insert to   people,places,
            things,events,info
produce/consume
                  information
                  systems
                  people,places,
                  things,events,info
prosume
          information
          systems
          people,places,
          things,events,info
mobile services
people,places,
things,events, info




                      information
                      systems
                      people,places,
                      things,events,info
social context
urban context
data-centric research
            preferences
        digital footprints
        social connections
information is
              abundant
     ubiquitously accessible
           heterogeneous
        widely distributed
an example
an example



             information is
                       abundant
              ubiquitously accessible
                    heterogeneous
                 widely distributed
an example
an example
x 600,000
how do we use the system?
x 600,000
how do we (group) use the system?
x 600,000
how do we (people) use the system?
x 600,000
what can we do with this data?
x 600,000
what can we do with this data?

address information abundance:
●   filter, personalise, rank
●   implicit behaviours to relevant,
    accurate, (timely) notifications
3. discussion/relevance
1. social computing techniques
are starting to be used to solve
software engineering tasks
2. can software engineering
techniques be used to solve
social computing tasks?
opportunities

●   we look at similar problems, in
    different contexts

●   we use similar terms, but with
    different interpretations

●   we adopt different solutions, that
    may be applicable
opportunities

●   data collection vs. requirements
    engineering

●   context-awareness vs. task-
    orientation

●   scalability (machine learning)
    vs. scalability (systems
    engineering)
Personalisation, Recommender Systems
SSE - CREST Workshop October 1, 2010
find me: @neal_lathia
http://www.cs.ucl.ac.uk/staff/n.lathia

CREST/SSE Workshop