Project #11
Title: Most Diverse Result set using Preference Queries
Leader's Name: Erik Ringdahl
Member2 Name: Johannes Moberg
Mamber3 Name: Carl Eriksson
Related paper: Orestis Gkorgkas, Akrivi Vlachou, Christos Doulkeridis and Kjetil Nørvåg. Finding the Most Diverse Products using Preference Queries.
Advances in Database Technology - EDBT 2015, 18th International Conference on Extending Database Technology, Brussels, Belgium, March 23-27, Proceedings. http://openproceedings.org/html/pages/2015_edbt.html
Presentation Day: May 25
Model: LE
Abstract: In this project we will, based on a set of users and database of products, examine the problem of returning the most diverse result set based on user preferences e.g. reaching the largest possible audience with an advertising campaign. We will evaluate the algorithms proposed by the authors and their usage in a commercial setting.
Queries on databases normally result in an exact match with a corresponding result set. However, exact matches are not always the optimal solution for many applications and scenarios. Top-k queries does not require exact matches, instead, such queries typically returns a set of the “top k” tuples that best matches the preferences of the user. These queries are often used to display a range of the most popular products from an overwhelmingly large database to a customer based on its preferences. Reverse top-k queries can be used to estimate the impact of a product on a market since it returns a set of preferences instead of a set of products. We will examine approaches to finding top-k sets and reverse top-k sets and apply the dispersion problem on the reverse sets.