Document ranking with sub-query series

• US 20080010268A1
• Filed: 07/06/2006
• Published: 01/10/2008
• Est. Priority Date: 07/06/2006
• Status: Active Grant

• Current Assignee: Oracle International Corporation (Oracle Corporation)
• Original Assignee: Oracle International Corporation (Oracle Corporation)
• Inventors: Liao, Ciya, Alpha, Shamim A.
• Granted Patent: US 7,849,077 B2
• Time in Patent Office: Days
• US Class Current: 707/5
• CPC Class Codes: G06F 16/3338 Query expansion
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

Systems, methods, media, and other embodiments associated with ranking documents by providing a search engine with a series of sub-queries generated from an original query are described. One example system includes input logic for receiving a query. The example system may include a relaxation logic configured to produce sub-queries from the query. The sub-queries may describe metadata string matching, content string matching, and/or metadata numerical attribute analysis. The sub-queries may be provided by an output logic to a search engine in an order that facilitates defining document relevance without requiring post-retrieval relevance ranking.

Claims

1. A computer implemented method, comprising:

receiving a query, the query being configured to cause a search logic to retrieve one or more documents, the documents comprising text elements and metadata elements;

creating from the query a series of sub-queries {SQ₁, . . . SQ_N}, N being an integer greater than 1, the series of sub-queries including one or more of, a metadata-centric sub-query, and a linguistic-centric sub-query,

the series of sub-queries being configured so that a sub-query SQ_X will cause the search logic to retrieve fewer documents than a sub-query SQ_Y when X<

Y;

the series of sub-queries being configured so that a sub-query SQ_A will cause the search logic to retrieve documents having a higher relevancy than a sub-query SQ_B when A<

B; and

providing, in order, one or more sub-queries from the series of sub-queries to the search logic until a pre-determined number of documents are retrieved by the search logic.

2. The method of claim 1, a metadata-centric sub-query comprising one or more of, a sub-query that matches terms in the query to one or more metadata elements on a phrase level, a sub-query that matches terms in the query to one or more metadata elements on a proximity level, and a sub-query that matches terms in the query to one or more metadata elements on a Boolean AND level.

3. The method of claim 2, a linguistic-centric sub-query including one or more of, a sub-query that matches terms in the query to one or more text elements on a phrase level, a sub-query that matches terms in the query to one or more text elements on a proximity level, and a sub-query that matches terms in the query to one or more text elements on a Boolean AND level.

4. The method of claim 3, including stemming terms in the query.

5. The method of claim 3, the series of sub-queries being ordered:

to make matches between terms in the query and metadata elements more determinative of document relevancy than matches between terms in the query and text elements;

to make a phrase match more determinative of document relevancy than a proximity match; and

to make a proximity match more determinative of document relevancy than a Boolean AND match.

6. The method of claim 5, the metadata elements including one or more of, a title, a description, a keyword, a subject, an author, a uniform resource locator (URL), an H1 headline, an H2 headline, and an anchor text.

7. The method of claim 6, a metadata element being associated with a metadata element weight, and a document relevance being a function of both:

(a) a type of match between terms in the query and a metadata element, and

(b) the metadata element weight.

8. The method of claim 7, including establishing the following metadata element to metadata element weight pairs:

(title, w₁), (description, w₂), (keyword, w₃), (subject, w₄), (author, w₅), (url, w₆), (H1 headline, w₇), (H₂ headline, w₈), and (anchortext, w₉), where w₁ through w₉ are configurable to facilitate determining document relevance before document retrieval.

9. The method of claim 8, where w₁>

w₂>

w₃>

w₄>

w₅>

w₆>

w₇>

w₈>

w₉.

10. The method of claim 5, the series of sub-queries including a sub-query based on one or more of, a numeric metadata link score, and a numeric metadata URL depth, where the numeric metadata link score is directly proportional to document relevance and where the numeric metadata URL depth is inversely proportional to document relevance.

11. A computer-readable medium storing processor executable instructions operable to perform a method, the method comprising:

receiving a query, the query being configured to cause a search logic to retrieve one or more documents, the documents comprising text elements and metadata elements;

creating from the query a series of sub-queries {SQ₁, . . . SQ_N}, N being an integer greater than 1, the series of sub-queries including one or more of, a metadata-centric sub-query, and a linguistic-centric sub-query; and

providing, in order, one or more sub-queries from the series of sub-queries to the search logic until a pre-determined number of documents are retrieved by the search logic.

12. A computer implemented method, comprising:

accessing an original request for documents;

producing from the original request for documents an ordered set of requests that when presented to a search engine will cause documents to be returned by the search engine in an order determined, at least in part, by document relevance; and

providing one or more members of the ordered set of requests to the search engine until a desired number of documents are returned,

a member of the ordered set of requests including one or more of, a metadata analysis based component, a linguistic analysis based component, and a link analysis based component, the metadata analysis based component, the linguistic analysis based component, and the link analysis based component being configured to facilitate determining document relevance before the search engine returns documents.

13. The method of claim 12, the metadata analysis based component associating a relevance r₁ with a phrase match, a relevance r₂ with a proximity match, and a relevance r₃ with a Boolean AND match, where r₁>

r₂>

r₃.

14. The method of claim 13, the metadata analysis based component associating a significance s₁ with a match with title metadata, a significance s₂ with a match with description metadata, a significance s₃ with a match with keyword metadata, a significance s₄ with a match with subject metadata, a significance s₅ with a match with author metadata, a significance s₆ with a match with uniform resource locator (URL) metadata, a significance s₇ with a match with H1 headline data, a significance s₈ with a match with H2 headline data, and a significance s₅ with an anchor text metadata, s₁ through s₉ being configurable to facilitate determining document relevance before the search engine returns documents.

15. The method of claim 14, the linguistic analysis based component associating a relevance r₄ with a phrase match, a relevance r₅ with a proximity match, and a relevance r₆ with a Boolean AND match, where r₄>

r₅>

r₆ and where r₁>

r₄.

16. The method of claim 15, the link analysis based component associating a relevance r₇ with a link score and a relevance r₈ with a URL depth, the link score being directly proportional to document relevance, the URL depth being inversely proportional to document relevance, r₇ and r₈ being configured to facilitate determining document relevance before the search engine returns documents.

17. The method of claim 16, including, upon determining that the original request has two or more query terms, producing the ordered set of requests to include, in order:

a request including phrase level matching between document metadata and the two or more terms, where the two or more terms may be stemmed, and also including one or more of, link score analysis, and URL depth analysis;

a request including proximity level matching between document metadata and the two or more terms, and also including one or more of, link score analysis and URL depth analysis;

a request including phrase level matching between document content and the two or more terms, where the two or more terms may be stemmed, and also including one or more of, link score analysis and URL depth analysis;

a request including proximity level matching between document content and the two or more terms, and also including one or more of, link score analysis, and URL depth analysis;

a request including proximity level matching between document content and the two or more terms; and

a request including AND level matching between document content and the two or more terms.

18. The method of claim 16, including upon determining that the original request has a single query term, producing the ordered set of requests to include, in order:

a request including matching between document metadata and the single term, and also including one or more of, link score analysis, and URL depth analysis;

a request including matching between document metadata and the single term, where the single term is stemmed, and also including one or more of, link score analysis, and URL depth analysis;

a request including matching between document content and the single term, and also including one or more of, link score analysis, and URL depth analysis; and

a request including matching between document content and the single term, where the single term is stemmed.

19. The method of claim 12, comprising:

receiving one or more documents from the search engine;

removing duplicate documents; and

presenting documents to a user.

20. A computer-implemented system, comprising:

an input logic configured to receive a query, the query being configured as an input for a search engine, the search engine being configured to retrieve documents in response to receiving a query, documents comprising text and metadata;

a relaxation logic configured to produce one or more sub-queries from the query, the sub-queries being configured to describe one or more of, metadata string matching, content string matching, and metadata numerical attribute analysis, the sub-queries being configured to define document relevance without performing post-retrieval relevance ranking; and

an output logic configured to provide one or more of the sub-queries to the search engine in an order that facilitates receiving relevant documents in order.

21. The system of claim 20, the relaxation logic being configured to produce one or more sub-queries from the query, the sub-queries being configured to include one or more of, phrase matching sub-queries, proximity matching sub-queries, and Boolean matching sub-queries.

22. The system of claim 21, where metadata available to match in a document includes one or more of, a title, a description, a keyword, a subject, an author, a URL, a headline, and an anchor text, and where metadata available to analyze numerically includes one or more of, a link score for a document, and a URL depth for a document.

23. The system of claim 22, the output logic being configured to provide sub-queries to the search engine until a desired number of documents have been retrieved.

24. The system of claim 23, comprising a presentation logic configured to provide retrieved documents to a user, the documents being ordered based on relevance without performing post-retrieval relevance ranking, the presentation logic being configured to selectively remove duplicate retrieved documents.

25. A system, comprising:

means for receiving a query intended for a search engine;

means for manipulating the query into one or more relaxed queries, the relaxed queries being configured to perform one or more of metadata matching, content matching, and metadata numerical analysis, the metadata matching and content matching being performed at one or more of a phrase level, a proximity level, and a Boolean level; and

means for providing one or more of the relaxed queries to the search engine in an order that pre-selects documents based on relevance.