A Bi-objective Approach for Composite Retrieval

MAURICIO MOYANO CASTILLO; PAULA ZABALA; GUILLERMO CABRERA-GUERRERO

Bologna

Conferencia; Joint EURO/ALIO International Conference 2018 on Applied Combinatorial Optimization; 2018

In information retrieval, traditional search strategies only consider one attribute of the elements to build up a ranking list of results which depends exclusively on the considered attribute. However, often one needs to rethink the original query to accomplish the right solution as these search strategies do not consider the existing relations among all the other attributes of the elements. As a response to this behavior, Composite Retrieval of Diverse and Complementary Bundles has been proposed. Its objective is to group elements into bundles, in which the items are related each other under both criteria: similarity and complementarity of bundles. In that way, the bundles should satisfy the users expectations without the needing of a new intervention, improving the searching experience. However, to the best of our knowledge,only single-objective models, mainly based on the weighted sum of each criterion, have been proposed in the literature. In this work we present a novel bi-objective Composite Retrieval model which considers the existing trade-of between diversity and complementarity of each set of bundles. A simple efficient local search algorithm is implemented to evaluate our model. We compare the obtained results with those from single objective models previously proposed in literature.