Cross-language information retrieval using meta-language index construction and structural queries
|Cross-language information retrieval using meta-language index construction and structural queries|
|Author(s)||Jadidinejad A.H., Mahmoudi F.|
|Published in||Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)|
|Keyword(s)||Unknown (Extra: Cross language information retrieval, Expert users, Ideal solutions, Index construction, Information need, Meta language, Precision and recall, State-of-the-art algorithms, Structural query, Structural query languages, Web search engines, Wikipedia, Computational linguistics, Information science, Query languages, Query processing, Search engines, Websites, Information retrieval)|
|Article||BASE, CiteSeerX, Google Scholar|
|Web||Ask, Bing, Google (PDF), Yahoo!|
|Download and mirrors|
|Local copy||Not available|
|Remote mirror(s)||Not available|
|Export and share|
|BibTeX, CSV, RDF, JSON|
|Browse properties · List of conference papers|
Cross-language information retrieval using meta-language index construction and structural queries is a 2010 conference paper written in English by Jadidinejad A.H., Mahmoudi F. and published in Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics).
Structural Query Language allows expert users to richly represent its information needs but unfortunately, the complexity of SQLs make them impractical in the Web search engines. Automatically detecting the concepts in an unstructured user's information need and generating a richly structured, multilingual equivalent query is an ideal solution. We utilize Wikipedia as a great concept repository and also some state of the art algorithms for extracting Wikipedia's concepts from the user's information need. This process is called "Query Wikification". Our experiments on the TEL corpus at CLEF2009 achieves +23% and +17% improvement in Mean Average Precision and Recall against the baseline. Our approach is unique in that, it does improve both precision and recall; two pans that often improving one, hurt the another.
- This section requires expansion. Please, help!
Probably, this publication is cited by others, but there are no articles available for them in WikiPapers. Cited 1 time(s)