PIX: a system for phrase matching in XML documents: a demonstration

Amer-Yahia, Sihem; Fernández, Mary; Srivastava, Divesh; Xu, Yu

doi:10.1109/icde.2003.1260864

Cited by 3 publications

(3 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Several use cases addressed in this paper have been studied in separate projects. For example, the PIX project at AT&T [4] has studied phrase search in the presence of comments and formatting instructions (i.e., the Excluded pattern described in Sect. 3).…”

Section: Related Workmentioning

confidence: 99%

“…Users who wish to search for all documents that 3. An f pattern to exclude certain fragments from an XML document for phrase search has been proposed in the PIX project [4]. Figure 5 gives an example.…”

Section: Excluded Patternmentioning

confidence: 99%

“…Figure 6d shows the rule that users can use in order to declare that they are interested to query the two instances separately. As for the Excluded pattern, the motivation for this pattern was provided by the PIX project [4]: Only the application of this pattern makes it possible to apply a phrase search for "Mickey likes Minnie", since the original document would not match. Both instances in Fig.…”

Section: Comment Patternmentioning

confidence: 99%

See 2 more Smart Citations

Predicate-based indexing for desktop search

2010

View full text Add to dashboard Cite

Google and other products have revolutionized the way we search for information. There are, however, still a number of research challenges. One challenge that arises specifically in desktop search is to exploit the structure and semantics of documents, as defined by the application program that generated the data (e.g., Word, Excel, or Outlook). The current generation of search products does not understand these structures and therefore often returns wrong results. This paper shows how today's search technology can be extended in order to take the specific semantics of certain structures into account. The key idea is to extend inverted file index structures with predicates which encode the circumstances under which certain keywords of a document become visible to a user. This paper provides a framework that allows to express the semantics of structures in documents and algorithms to construct enhanced, predicate-based indexes. Furthermore, this paper shows how keyword and phrase queries can be processed efficiently on such enhanced indexes. It is shown that the proposed approach has superior retrieval performance with regard to both recall and precision and has tolerable space and query running time overheads.

show abstract

Section: Related Workmentioning

confidence: 99%

Section: Excluded Patternmentioning

confidence: 99%

Section: Comment Patternmentioning

confidence: 99%

See 1 more Smart Citation