Context-sensitive wordless search

• US 20040059726A1
• Filed: 09/09/2003
• Published: 03/25/2004
• Est. Priority Date: 09/09/2002
• Status: Abandoned Application

• Current Assignee: Relevate Software, Inc.
• Original Assignee: Relevate Software, Inc.
• Inventors: Hunter, Jeff, Padmanabhan, Ranjit
• Application Number:
  US10/659,557
• Publication Number:
  US 20040059726A1
• Time in Patent Office: Days
• US Class Current: 707/3
• CPC Class Codes: G06F 16/3325 Reformulation based on results of preceding query

Abstract

Embodiments of the invention provide the searcher a unique method of interacting with one or more documents to build a context-sensitive query that can retrieve additional documents that are closer to the searcher'"'"'s needs. Embodiments of the invention do not require the searcher to enter any text and translate the searcher'"'"'s intent into complex queries that will be executed by existing search engines. Embodiments of the invention iteratively modify the context-sensitive query and eventually retrieve a document that satisfies the searcher'"'"'s requirements.

Claims

1. A method of performing a context-sensitive search comprising:

accepting a selection of a first document;

accepting a selection of a first term from within the first document;

determining a context of the first term with respect to the first document;

choosing at least two documents that contain the first term; and

ranking the at least two documents that contain the first term according to how closely a context of the first term with respect to the at least two documents matches the context of the first term with respect to the first document.

2. The method of claim 1, wherein accepting a selection of a first term from within the first document comprises:

accepting a selection of the first term in response to a device chosen from the group consisting of a computer mouse, a trackball, a joystick, a touchpad, and a laser pointer.

3. The method of claim 1, wherein accepting a selection of a first term from within the first document comprises:

accepting a selection of the first term in response to a sound.

4. The method of claim 1, further comprising:

accepting a selection of a second term from the first document;

determining a context of the second term with respect to the first document;

associating a first modifier that is indicative of the relevancy of the first term with the first term;

associating a second modifier that is indicative of the relevancy of the second term with the second term;

instead of choosing at least two documents that contain the first term, choosing at least two documents that contain the first and second terms; and

ranking the at least two documents that contain the first and second terms according to how closely a context of the first and second terms with respect to the at least two documents matches the context of the first and second terms with respect to the first document, and according to the first and second modifiers.

5. The method of claim 4, wherein determining a context of the first term with respect to the first document and determining a context of the second term with respect to the first document comprises:

identifying whether any structural tags exist in the first document.

6. The method of claim 5, wherein identifying whether any structural tags exist in the first document comprises:

determining whether the first document is characterized as one belonging to a group consisting of a document with no structural tags and no discernible structure, a document with no structural tags and a discernible structure, a document with a structural tag that has physical markup, and a document with a structural tag that has physical and logical markup.

7. The method of claim 6, wherein a document with a structural tag that has physical markup comprises a HTML document.

8. The method of claim 6, wherein a document with a structural tag that has physical and logical markup comprises a document that complies with an XML schema.

9. The method of claim 4, further comprising:

accepting a selection of a third term from one of the at least two documents that contain the first and second terms;

determining a context of the third term with respect to the one of the at least two documents that contain the first and second terms;

assigning a third modifier to the third term based upon the relevancy of the third term;

choosing at least two documents that contain the first, second, and third terms; and

ranking the at least two documents that contain the first, second, and third terms according to how closely a context of the first and second terms with respect to the at least two documents that contain the first, second, and third terms matches the context of the first and second terms with respect to the first document, according to how closely a context of the third term with respect to the at least two documents that contain the first, second, and third terms matches the context of the third term with respect to the one of the at least two documents that contain the first and second terms, and according to the first, second, and third modifiers.

10. The method of claim 4, wherein associating a first modifier with the first term and associating a second modifier with the second term comprise:

associating a modifier with the first term and with the second term that is chosen from the group consisting of more relevant, less relevant, not relevant, and exactly relevant.

11. A method comprising:

assigning a first document a complexity rating that is indicative of the complexity of the first document'"'"'s structure;

associating a relevance indicator with a first element that is contained within the first document; and

finding a second document based upon the second document'"'"'s complexity rating being no greater than the first document'"'"'s complexity rating, based upon a relationship between the first element and the first document being the same as a relationship between a second element in the second document and the second document, and based upon the similarity between the first element and the second element.

12. The method of claim 11, wherein finding the second document additionally comprises:

constructing a query; and

sending the query to a search engine that uses the query to find the second document.

13. The method of claim 11, wherein associating the relevancy indicator with the first element comprises accepting an input in response to a device that performs a highlighting function.

14. The method of claim 11, wherein associating the relevancy indicator with the first element comprises assigning a less relevant indicator to the first element.

15. The method of claim 11, wherein associating the relevancy indicator with the first element comprises assigning a more relevant indicator to the first element.

16. The method of claim 11, wherein assigning the first document a complexity rating that is indicative of the complexity of the first document'"'"'s structure comprises:

assigning the first document a first rating if the first document has no structural tags and no discernible structure;

assigning the first document a second rating if the first document has no structural tags but a discernible structural pattern;

assigning the first document a third rating if the first document has structural tags with physical markup; and

assigning the first document a fourth rating if the first document has structural tags with physical and logical markup.

17. The method of claim 12, further comprising:

associating a relevance indicator with a second element that is contained within the second document; and

modifying the query by incorporating the second element and its relevance indicator.

18. A device-readable medium that, when read, causes a first device to perform processes comprising:

storing a file that contains structural information about a document;

storing at least one fragment from the document in response to a first external input;

storing a modifier that indicates the relevancy of the at least one fragment in response to a second external input;

forming a context-sensitive search query based upon the modifier, the at least one fragment, and the file;

sending the context-sensitive search query to a second device to find a first plurality of result set items that conforms to the context-sensitive search query.

19. The medium of claim 18, where analyzing the structure of the document further comprises:

determining whether the document has logical markup data, physical markup data, and an observable structural pattern.

20. The medium of claim 18, further causing the first device to perform processes further comprising:

storing a result set item fragment from one of the plurality of result set items in response to a third external input;

storing another modifier that indicates the relevancy of the result set item fragment in response to a fourth external input;

forming a modified context-sensitive query based upon the result set item fragment and the another modifier; and

sending the modified context-sensitive search query to the second device that finds a second plurality of result set items conforming to the modified context-sensitive search query.

21. A method of performing a context-sensitive search comprising:

under control of a client system,

displaying a document;

associating a text fragment in the document with a modifier based on inputs from a searcher;

sending a request to find other documents that contain the text fragment to a server system; and

under control of the server system,

receiving the request;

building a query that is responsive to the context of the text fragment in the document and that is also responsive to the modifier; and

submitting the query to a search engine.

22. The method of claim 21, wherein the server system additionally:

receives results from the search engine; and

sends the received results to the client system.