Main Research Themes

  • Harnessing and Advancing Social Search (HASS)
  • Models for the Collaborative Web
  • Understanding Email Search and Re-finding
  • VAMOS Healthcare Project
  • Casual Leisure Search

  • Harnessing and Advancing Social Search (HASS):

    Understanding User Intent, Information Need and Temporal Relevance

    We propose to investigate searching and browsing in social sites and determine where the one-shot and one-size-fits-all paradigm of search is failing users and does not sufficiently assist them with their information gathering task. We will use modern statistical learning techniques to develop models that are able to utilise personalisation, temporal task-based knowledge and topical information derived from the corpus to improve search. The proposed work will significantly extend earlier work in personalisation of social media search and latent topic models carried out by the applicant. These new models will better serve users their information needs and better support them in completing more complex tasks over multiple queries or even sessions. Furthermore the models will provide better insight into the data contained in social sites including information about the topics represented and how their use and popularity is varying over time.

    Most relevant publication (more):

    Harvey, M., Crestani, F., Carman, M.
    Building User Profiles from Topic Models for Personalised Search
    ACM 22nd Conference on Information and Knowledge Management, CIKM 2013. San Francisco, CA, USA. (October 2013, acceptance rate: 12.5%)
    Conference web site

    Models for the Collaborative Web

    Since its creation in the early 1990s the Web has held the promise of allowing near-instantaneous communication, participation and sharing of resources and ideas between users across the globe. However up until fairly recently, the Web was predominantly a large collection of static documents providing no real scope for such interaction. The past decade has seen the arrival and rapid growth of the so called "Web 2.0"€¯ movement where sites have become increasingly more social with users able to share information with others.
    This work explores two avenues of this new social web: social tagging and ratings-based collaborative filtering and introduces a family of novel latent variable Bayesian models designed for this data. A series of experiments carried out on real-world data sets show that these models can overcome the inherent difficulties and provide significant improvements in performance over state of the art systems. Furthermore it is shown that the output of these models is more readily interpretable than from competing models and can therefore be utilised to gain a more complete understanding of the complex social and topical dynamics of such systems.

    Most relevant publication (more):

    Harvey, M.
    Bayesian Latent Variable Models for the Collaborative Web
    PhD Thesis, Strathcyde University. Glasgow, Scotland. (May 2011)
    Thesis (PDF)

    Understanding Email Search and Re-finding

    Despite Email being the most popular communication medium currently in use and the fact that people have been shown to regularly re-use messages, very little is known about how people actually search within email clients. In this work we present a detailed analysis of email search behaviour and uncover a number of behavioral patterns that contrast with those previously observed in web search. From these findings, we describe ways in which email search could be improved in future.

    Most relevant publications (more):

    Harvey, M., Elsweiler, D.
    Exploring Query Patterns in Email Search
    In Proceedings Advances in Information Retrieval, 34th European Conference on IR Research, ECIR 2012. Barcelona, Spain. (April 2012, acceptance rate: 21%)
    Conference web site

    Elsweiler, D., Harvey, M., Hacker, M.
    Understanding Re-finding Behaviour in Naturalistic Email Interaction Logs
    Proceedings of the 34th Annual ACM SIGIR Conference. Beijing, China. (July 2011, acceptance rate: 19.8%) **honourable mention**
    Conference web site

    VAMOS Healthcare Project

    The main aims of the VAMOS project is to expand the supply and service opportunities in healthcare in the home through the use of the latest assistive technologies and continue to simultaneously measure the achieved quality. Related work involves the automatic recommendation of meal plans to users that will suit both their tastes and nutritional needs.

    In order to achieve these goals it is neccesary to take a number of incremental steps:

  • accurately measure the activity levels of users and then use these measurements to model nutritive requirements
  • build models which can accurately learn user food preferences
  • finally, using both of these sources of knowledge, automatically generate meals plans that optimally fit both a user's nutritional requirements and his/her tastes

  • Most relevant publications (more):

    Harvey, M., Elsweiler, D., Ludwig, B.
    You are what you eat: learning user tastes for rating prediction
    20th String Processing and Information Retrieval Symposium (SPIRE). Jerusalem, Israel. (October 2013)
    Conference web site

    Müller, M., Harvey, M., Elsweiler, D., Mika, S.
    Ingredient Matching to Determine the Nutritional Properties of Internet-Sourced Recipes
    In Proceedings Pervasive Health 2012, San Diego, California. (May 2012, acceptance rate: 34%)
    Conference web site

    Casual Leisure Search

    In this project we seek to investigate how people can be assisted when out and about visiting multiple events over a fixed amount of time. We seek to both recommend events of interest to people and also to automatically generate tours allowing them to visit as many interesting events as possible. Through analysis of log files, surveys and questionnaires we learn how people search in casual leisure settings where they are not searching to meet a particular information need but rather of entertainment purposes.

    I am serving as an organiser for an up-coming workshop on Casual Leisure Search called "Search 4 Fun" which will be co-located with ECIR2012 in Barcelona. For more information please visit the workshop web site.

    Most relevant publications (more):

    Schaller, R., Harvey, M., Elsweiler, D.
    Entertainment on the Go: Finding Things to Do and See while Visiting Distributed Events
    In Proceedings Fourth Information Interaction in Context Symposium (IIiX 2012), Nijmegen, the Netherlands. (August 2012)
    Conference web site

    Schaller, R., Harvey, M., Elsweiler, D.
    Out and About on Museums Night: Investigating Mobile Search Behaviour for Leisure Events
    In Proceedings Searching4Fun Workshop (S4F 2012), ECIR 2012 Barcelona, Spain. (April 2012)
    Conference web site