Projects

Posted on Author Konstantinos Stefanidis

Here you can find some recommendations for your project topic. These topics can act as examples. Feel free to propose your own topic!

  • Sequential recommendations: Recommendations should not be considered as a stand-alone process. A real world system requires access to history logs, i.e., to know what items the system previously had recommended, how well were these items received by the users, so as to take these factors into account when producing recommendations.
  • Sequential fair recommendations for groups: Use the notion of satisfaction, that describes how relevant are the recommended items to each member of the group, to ensure high satisfaction for all group members.
  • Interactive sequential (group) recommendations: Users provide feedback of several forms to the system, which is used for improving the quality of recommendations. 
  • Visualisation / Explanation of sequential recommendations: I suggest “A”, which is “excellent” for “X”, because he/she was not satisfied in the previous round of suggestions. 
  • Chart-like explanations / visualisations for recommendations: I suggest “A” because 90% of your “friends”, aka. similar user, like it. 
  • Natural Language Explanations for Recommendations
  • Explaining recommendation via why-not queries: Understand why certain items are no recommended. For example, an explanation for not obtaining Titanic as recommended movie could be a very low rating of a very similar user. In this project, we are aiming to formalise the problem of Why-Not queries in recommendation systems, and propose ways to compute them, either in the source data or the filtering function itself.
  • Recommendation summaries: Use the notion of coverage or diversity to report representative recommendations. 
  • Recommendation summaries: Use the notion of coverage or diversity to report textual summaries of recommendations. 
  • Education and emotion based group recommendations for health: Existing systems recommend to groups of persons health documents selected by caregivers, by incorporating the notion of fairness. This project focuses on adapting recommendations considering the educational level of the end-users and their psycho-emotional status. 
  • Recommend product packets to customers: E.g., a mobile phone along with its case and headphones. 
  • Recommendations by example: Introduce a novel paradigm that considers a user query as an example of the data in which the user is interested. 
  • Recommend (representative) reviews to users 
  • Recommendations based on User Reviews
  • Recommendation on a map: E.g., hotels 
  • Survey on sequential recommendations 
  • Survey on example-driven search 
  • Survey on fairness in recommender systems

Projects examples from the previous year: