Dynamic strategies for query constructing and rank merging from multiple search engines

Heterogeneous search engines differ in the algorithms they use and the domains they cover, thus there is no single search engine that performs best in every circumstance.

In order to obtain optimal search results, it often makes sense to use more than one search engine. However, appropriately merging results from different engines is challenging, i.e. combining results in such a way that they reflect the ranking of results the user would choose. In this paper, we propose an effective way to achieve this for web services search which can be extended to cloud services and be applied to big data. In contrast to ”classical” search processed by conventional text-based search engines, a more elaborated search request is needed here. In addition to the result merging, we therefore present a method to create a structured request for this specific task. The evaluation of our proposed solution shows that it is satisfying in terms of both result quality and performance.

@inproceedings{DBLP:conf/esocc/OpasjumruskitKE15,
  author    = {Kobkaew Opasjumruskit and
               Birgitta K{\"{o}}nig{-}Ries and
               Jes{\'{u}}s Exp{\'{o}}sito},
  title     = {Dynamic Strategies for Query Constructing and Rank Merging from Multiple
               Search Engines},
  booktitle = {Service Oriented and Cloud Computing - 4th European Conference, {ESOCC}
               2015, Taormina, Italy, September 15-17, 2015. Proceedings},
  pages     = {141--155},
  year      = {2015},
  crossref  = {DBLP:conf/esocc/2015},
  url       = {http://dx.doi.org/10.1007/978-3-319-24072-5_10},
  doi       = {10.1007/978-3-319-24072-5_10},
  timestamp = {Tue, 08 Sep 2015 13:41:37 +0200},
  biburl    = {http://dblp.dagstuhl.de/rec/bib/conf/esocc/OpasjumruskitKE15},
  bibsource = {dblp computer science bibliography, http://dblp.org}
}