Section: New Results

Distributed Indexing and Searching

We propose a new diversified search and recommendation solution suited for scientific data (i.e., plant phenotyping, botanical data) [22] We first define an original profile diversification scoring function that enables to address the problem of returning redundant items, and enhances the quality of diversification. Through experimental evaluation using two benchmarks, we showed that our scoring function gives the best compromise between diversity and relevancy. Next, to implement our new scoring function, we propose a basic Top-k threshold-based algorithm that exploits a candidate list to achieve diversification and several techniques to improve performance. First, we simplify the scoring model to reduce its computational complexity. Second, we propose two techniques to reduce the number of items in the candidate list, and thus the number of diversified scores to compute. Third, we propose different indexing scores that take into account the diversification of items and an adaptive indexing approach to reduce the number of accesses in the index dynamically based on the queries workload. The experimentation results show that our techniques yield a major reduction of response time, up to 12 times compared to a baseline greedy diversification algorithm.

We also address distributed and diversified recommendation in the context of P2P and multisite cloud [23] . We propose a new scoring function (usefulness) to cluster relevant users over a distributed overlay. Our experimental evaluation using different datasets shows major gains in recall (order of 3 times) compared with state-of-the-art solutions.