Ellen Vorhees. "Multiple Search Engines in Database Merging", 1997. dl 97. 7  Home/Search Document Details and Download
Summary TOC Related Articles Check |
This paper is cited in the following contexts:
Methodologies for Distributed Information Retrieval - de Kretser, Moffat, Shimmin, .. (1998) (3 citations) (Correct)
....and receptionists. In the alternative federated methodologies described below we assume that the librarians and receptionist are similar enough to share information such as vocabulary and index and use the same similarity heuristic, but other arrangements are possible. For example, Voorhees et al. [21, 22, 23, 24] have described strategies for merging the rankings returned by librarians with no knowledge of how they were computed, so long as appropriate training queries are available. Evaluation Criteria To evaluate performance we need to consider three factors. One is e#ectiveness . Another is response ....
E.M. Voorhees and R.M. Tong. Multiple search engines in database merging. In R.B. Allen and E. Rasmussen, editors, Proc. ACM Digital Libraries, pages 93--102, Philadephia, Pennsylvania, 1997.
Using navigation data to improve IR functions in the context.. - Hansen, Shriver (Correct)
....the similarity of their associated search sessions. In turn, by combining search sessions with queries in a given group, we can better identify relevant URLs. In the IR literature, there are several examples in which the pattern of retrieved items is used to form so called query clusters, see [16, 17]. In our context, these schemes would involve the queries submitted by users together with the top L relevant pages returned by a given search engine. Our technique is different in that we consider only those pages that were actually selected by a user during a search task. This has the effect of ....
E. M. Voorhees and R. M. Tong. Multiple search engines in database merging. In Proceedings of the Second ACM International Conference on Digital Libraries, pages 93--102, Philadelphia, PA, July 1997.
Machine Translation for Information Access across the . . . - Lin (1999) (Correct)
....a large semantic knowledge base by ontology alignment, dictionary parsing, and web mining to overcome the meaning In Machine Translation Summit VII, September 13 17, 1999, Singapore. Page 9 of 9 fanout problem. Initial results are reported in Lin Hovy [16] and Hovy [10] Voorhees Tong [23] report that fusing retrieval results from multiple collections could achieve better performance than from a single collection. MuST currently directs users queries to a single database. Allowing MuST users to a submit single query and search all the available collections would be a good addition ....
Voorhees, E. M., Tong, R. M. Multiple Search Engines in Database Merging. In Proc. of the 2 nd ACM Int'l Conf. on Digital Libraries, pp. 93-102, 1997.
Democratic Data Fusion for Information Retrieval Mediators - Tzitzikas (2001) (Correct)
....systems, is certainly an ill founded approach, and may result to low retrieval e#ectiveness. For instance, the approach for server selection proposed in [4] presupposes that the mediator knows the number of relevant documents in each underlying system In some other approaches (i.e. 12] [13]) they exploit the results of past (or training) queries for estimating the number of relevant documents of each underlying system. However even this approach goes against the autonomy and continuous evolution of the systems of the web. However, according to our view, the availability of more ....
....approach is better founded and the technique presented in this paper is based on this hypothesis. Let us now focus on the problem of results fusing. Some approaches assume that the degrees of relevance returned by each system are comparable, and they use them for ordering the results (i.e. 12] [13]) while some others (i.e. 4] just interleave the returned orderings. In [2] two isolated techniques for merging the search results are introduced. These techniques require downloading the document contents and they also employ a set of relevance collection statistics The drawbacks of these ....
Ellen Vorhees. "Multiple Search Engines in Database Merging", 1997. dl 97. 7
Query-Based Sampling of Text Databases - Callan, Connell (1999) (23 citations) (Correct)
....These two algorithms represent databases by their prior e ectiveness for past queries, which makes it easy to control their behavior relatively precisely without knowing anything about the contents of the database. They also make it easy to integrate databases served by di erent search engines [37], because there is no need to compare representations or frequencies produced by di erent search engines. The main problem in applying these algorithms is that relevance judgements require manual e ort, so it can be expensive to apply them when there are many databases or databases that are ....
E.M. Voorhees and R.M. Tong. Multiple search engines in database merging. In Proceedings of the 2nd ACM International Conference on Digital Libraries, Philadelphia, 1997. ACM.
Collection Selection and Results Merging with Topically.. - Larkey, Connell, Callan (2000) (7 citations) (Correct)
....be selected. Some researchers rely on manually created characterizations of the collections [4] others require a set of reference queries or topics with relevance judgements, and select those collections with the largest numbers of relevant documents for topics that are similar to the new query [17]. We are interested in the class of approaches including CORI [1] gGlOSS [6] and others [8] 20] that characterize different collections using collection statistics like term frequencies. These statistics, which are used to select or rank the available collections relevance to a query, are ....
Voorhees, E. M. and Tong, R. M. Multiple Search Engines in Database Merging. In Digital Libraries 97, The 2 nd ACM International Conference on Digital Libraries, Philadelphia. pages 93-102, 1997.
Mediating and Metasearching on the Internet - Gravano, Papakonstantinou (1998) (4 citations) (Correct)
....metasearcher s scores. An approach to address these problems is to learn from the results of training queries. Given a new query, the closest training queries are used to determine how many documents to extract from each available collection, and how to interleave them into a single document rank [29, 30]. Another approach is to calibrate the document scores from each collection using statistics about the word distribution in the collections [1] The STARTS protocol asks sources to report term statistics in their query results. For example, the entry for a document d in the result for a query ....
E. M. Voorhees and R. M. Tong. Multiple search engines in database merging. In Proceedings of the Second ACM International Conference on Digital Libraries (DL'97), July 1997.
Democratic Data Fusion for Information Retrieval Mediators - Yannis Tzitzikas Department (2001) (Correct)
No context found.
Ellen Vorhees. "Multiple Search Engines in Database Merging", 1997. dl 97. 7
Democratic Data Fusion for Information Retrieval Mediators - Yannis Tzitzikas Department (2001) (Correct)
No context found.
Ellen Vorhees. "Multiple Search Engines in Database Merging", 1997. dl 97. 7
Web Search Services - Jiying Wang And (Correct)
No context found.
Voorhees, E.M., and Tong, R.M., "Multiple search engines in database merging," Proc. 2 ACM Intl. Conf. on Digital Libraries, 93-102, 1997. Available at http://www-nlpir.nist.gov/works/papers/dl97.ps
Democratic Data Fusion for Information Retrieval Mediators - Yannis Tzitzikas Department (2001) (Correct)
No context found.
Ellen Vorhees. "Multiple Search Engines in Database Merging", 1997. dl 97. 7
Building Efficient and Effective Metasearch Engines - Meng, Yu, Liu (2002) (11 citations) (Correct)
No context found.
E. Voorhees, and R. Tong. Multiple Search Engines in Database Merging. Second ACM International Conference on Digital Libraries, July 1997.
Online articles have much greater impact More about CiteSeer.IST Add search form to your site Submit documents Feedback
CiteSeer.IST - Copyright Penn State and NEC