Improving the efficiency of distributed information retrieval using hybrid index partitioning

Download
2018
Hafızoğlu, Fatih
Selective search with traditional partitioning have advantages over exhaustive search in terms of total query cost. However, it can suffer from query latency and load imbalance for most of the time due to its nature. To overcome these issues, we proposed a new partitioning method in this thesis, namely Hybrid partitioning. Our studies shows that it is possible to obtain significant savings in query latency with this new partitioning methodology. In addition to that, query processing with Hybrid partitioning also achieves perfect load balancing and provides resource optimization, which is a key point for low resource environments.
Citation Formats
F. Hafızoğlu, “Improving the efficiency of distributed information retrieval using hybrid index partitioning,” M.S. - Master of Science, Middle East Technical University, 2018.