Presentation of structured search results

• US 7,792,821 B2
• Filed: 06/29/2006
• Issued: 09/07/2010
• Est. Priority Date: 06/29/2006
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Shakib, Darren A., Flake, Gary William, Naam, Ramez
• Primary Examiner(s): Filipczyk; Marc R
• Application Number:
  US11/427,693
• Publication Number:
  US 20080005118A1
• Time in Patent Office: 1,531 Days
• Field of Search: 707 1- 6, 707/10, 707/100, 707/101, 707/714, 707/749, 707999003-999006, 707/999.01
• US Class Current: 707/714
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access

Abstract

A method and system is provided for automatically displaying a search-refinement user interface based on structure identified in search results in response to unstructured search queries. A list of search results obtained by a search engine is analyzed to determined whether a structure threshold is satisfied. If the threshold is met, then search-refinement options are displayed on a common display with the search results to allow the user to filter the search results according to the structure identified in the results. The user interface can receive filter and sort requests from the user to further refine the search results based on overlapping schema attributes or other shared-structure elements identified in the search results.

Claims

1. A computer-implemented method of automatically providing search query refinement options using a computing device having processor, memory, and data storage subsystems, the computer-implemented method comprising:

receiving an unstructured web-search query request from a user via a user input component;

initiating an unstructured web-search query according to the unstructured web-search query request;

obtaining a first set of web-search results in response to said unstructured web-search query, wherein the first set of web-search results are ranked in order of relevance;

determining whether sufficient common structure attributes are present in the first set of web search results to satisfy a structure threshold, wherein the structure threshold is satisfied if a sufficient percentage or absolute count of the first set of web search results have one or more common structure attributes, and the common structure attributes comprise similar name-value pairs having shared common name fields in meta-tags or in XML-code contained in the first set of search results;

analyzing a subset of top-ranked web-search results from the first set to identify a common structure attribute present in the subset;

analyzing the first set of web-search results according to the identified common structure attribute in the subset analysis;

automatically generating a set of search refinement options according to the analysis of the first set of web-search results according to the identified common structure attribute;

communicating the set of search refinement options for display to a user output device;

receiving a selection of one of the search refinement options; and

communicating a second set of web-search results resulting from the selected search refinement option to the user output device.

2. The computer-implemented method of claim 1, further comprising:

filtering the subset of top-ranked web-search results according to the common structure attribute identified.

3. The computer-implemented method of claim 2, further comprising:

automatically communicating one or more columns listing, for each search result, a value associated with each selected common structure attribute.

4. The computer-implemented method of claim 1, wherein a plurality of common structure attributes comprise overlapping schema attributes.

5. One or more computer storage media having computer-executable instructions embodied thereon that, when executed by a computing device, perform a method of providing search query refinement options, the method comprising:

receiving an unstructured web-search query request from a user via a user input component;

initiating an unstructured web-search query according to the unstructured web-search query request;

obtaining original web-search results in response to said unstructured web-search query, wherein the original web-search results are ranked in order of relevance;

determining whether sufficient common structure attributes are present in the original web search results to satisfy a structure threshold, wherein the structure threshold is satisfied if a sufficient percentage or absolute count of the original web search results have one or more common structure attributes, and wherein the common structure attributes comprise similar name-value pairs having shared common name fields in meta-tags or in XML-code contained in the original web search results;

obtaining a highly relevant subset of results from the ranked original web-search results;

analyzing said highly relevant subset of results to extract one or more common attributes present in the highly relevant subset of results;

analyzing the original web-search results to generate one or more search filters based upon the one or more common attributes present in the highly relevant subset of results;

communicating a user option for selecting one of the generated search filters to a user structured search results interface;

refining the original web-search results based upon the selected generated search filter; and

returning the original web-search results and the refined original web-search results based upon the selected generated search filter, to the user structured search results interface.

6. The method of claim 5, wherein said original web search results comprise internet-search results and said web search query comprises an internet-search query.

7. The method of claim 5, further comprising determining whether said original web search results vary sufficiently to satisfy a variance threshold with respect to a particular overlapping schema attribute from a plurality of overlapping schema attributes.

8. The method of claim 7, wherein one or more of said plurality of overlapping schema attributes are common to a shared formal schema.

9. The method of claim 5, further comprising receiving a request to refine said web search results by applying a filter based on at least one selected overlapping schema attribute.

10. In a computer system having a processor, memory and data storage subsystems, a computer-implemented system for automatically providing a structured-search-results user interface in response to unstructured queries, the computer-implemented system comprising:

a user input component, to receive an unstructured search query request from a user;

a data store comprising information elements, wherein the information elements comprise structured data, semi-structured data, and unstructured data;

a search engine component to obtain original search results from the data store in response to the unstructured search query request wherein the original search results are ranked in order of relevance, and further the search engine component to obtain a highly relevant subset of results from the ranked original search results;

an analytical component to determine whether sufficient common structure attributes are present in the original search results to satisfy a structure threshold, wherein the structure threshold is satisfied if a sufficient percentage or absolute count of the original search results have one or more common structure attributes which comprise similar name-value pairs having shared common name fields in meta-tags or in XML-code contained in the original search results, and the analytical component to further determine one or more categories present in the highly relevant subset of results and to generate a search filter for the original search results based upon a variance threshold of the one or more categories present in the highly relevant subset of results; and

an interface component to communicate the original search results, a user option for selecting the generated search filter from an analysis of the highly relevant subset of results, and refined original search results based upon the selected generated search filter via a request from the user input component.

11. The computer-implemented system of claim 10, wherein the original search results comprise internet-search results and the search query comprises an internet-search query.

12. The computer-implemented system of claim 11, wherein the search engine component further determines whether the original search results vary sufficiently to satisfy a variance threshold with respect to a particular overlapping schema attribute from a plurality of overlapping schema attributes.

13. The computer-implemented system of claim 12, wherein the plurality of overlapping schema attributes are common to a shared formal schema, and further wherein the communicated generated search filter comprises filter and sort options specified by the shared formal schema.

14. The computer-implemented system of claim 10, wherein the interface component receives a request to refine the original search results by applying a filter based on at least one selected overlapping schema attribute.

15. The computer-implemented system of claim 14, wherein the interface component automatically communicates one or more columns listing, for each original search result, a value associated with each selected category.