Method and system of search result diversification

• US 8,583,640 B2
• Filed: 10/04/2010
• Issued: 11/12/2013
• Est. Priority Date: 11/12/2009
• Status: Active Grant

• Current Assignee: Alibaba Group Holding Ltd.
• Original Assignee: Alibaba Group Holding Ltd.
• Inventors: Zhang, Li, Li, Jinhui, Guo, Ruijie
• Primary Examiner(s): LE, DEBBIE M
• Application Number:
  US12/992,399
• Publication Number:
  US 20120221562A1
• Time in Patent Office: 1,135 Days
• Field of Search: 707/722, 707/706, 707/726, 707/748, 707/707, 707/728, 707/752
• US Class Current: 707/728
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/248 Presentation of query results
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques

Abstract

The present disclosure discloses a search method and system. A method obtains a first search result set of first search results relevant to query data submitted by a client. According to a first relevance score and a preset diversity field of each first search result in the first search result set, a second relevance score of each first search result is calculated. The preset diversity field represents an attribute category of a respective first search result. According to the first relevance score and the second relevance score, a relevance parameter score for each first search result is generated. According to a preset number of second search results and the relevance parameter score, the present number of second search results are extracted from the first search result set to display to the client. The technique achieves lower consumption of system resources, faster computation speed and more flexibility in diversification of search results.

Claims

1. A search method comprising:

under control of a computer configured with executable instructions;

according to query data submitted by a client, obtaining a first search result set of first search results relevant to the query data;

according to a first relevance score and a preset diversity field of each first search result in the first search result set, calculating a second relevance score of each first search result, the preset diversity field representing an attribute category of a respective first search result;

according to the first relevance score and the second relevance score of each first search result, generating a relevance parameter score for each first search result, wherein the generating comprises summing the first relevance score and the second relevance score to provide the relevance parameter score for each first search result; and

according to a preset number of second search results and the relevance parameter score of each first search result, extracting the preset number of second search results from the first search result set for display to the client, wherein the extracting comprises;

classifying the first search result set to obtain a plurality of subsets;

according to the relevance parameter score, ranking each subset of the plurality of subsets after classification of the first search results; and

according to a ranking order, extracting the preset number of second search results respectively from the ranked subsets, the preset number of second search results being a product of a number of diversity values and a number of recurring extractions.

2. The method as recited in claim 1, wherein classifying the first search result set to obtain a plurality of subsets comprises:

according to the preset diversity field of each first search result in the first search result set, classifying the first search result set to obtain a respective subset corresponding to each respective attribute category of the first search result set, and calculating the second relevance score of each first search result comprises;

according to the first relevance score in each subset, obtaining a corresponding ranking position of a respective first search result; and

according to a preset relationship between the ranking position of the respective first search result and the second relevance score, obtaining the second relevance score of each first search result.

3. The method as recited in claim 1, further comprising:

storing the query data, the preset number of second search results, and a corresponding relationship between the query data and the preset number of second search results into a database.

4. The method as recited in claim 1, wherein obtaining the first search result set of first search results relevant to the query data comprises:

according to the first relevance score, conducting a search based on the query data submitted by the client; and

according to the preset diversity field, extracting the first search results from search results of the search.

5. The method as recited in claim 1, further comprising:

displaying the preset number of second search results to the client.

6. A search system comprising:

one or more processors;

storage media;

a retrieval unit stored in the storage media and executable by the one or more processors that, according to query data submitted by a client, obtains a first search result set of first search results relevant to the query data;

a calculation unit stored in the storage media and executable by the one or more processors that, according to a first relevance score and a preset diversity field of each first search result in the first search result set, calculates a second relevance score of each first search result, the preset diversity field representing an attribute category of a respective first search result;

a configuration unit stored in the storage media and executable by the one or more processors that, according to the first relevance score and the second relevance score, generates a relevance parameter score of each first search result, wherein the configuration unit sums the first relevance score and the second relevance score to provide the relevance parameter score of each first search result; and

an extraction unit stored in the storage media and executable by the one or more processors that, according to a preset number of second search results and the relevance parameter score, extracts the present number of second search results from the first search result set to display to the client, wherein the extraction unit comprises;

a ranking sub-unit that, according to the relevance parameter score, ranks a plurality of subsets of the first search result set that are obtained after classification of the first search results; and

a first extraction sub-unit that, according to a ranking order, extracts the preset number of second search results from the ranked subsets, the preset number of second search results being a product of a number of diversity values and a number of recurring extractions.

7. The system as recited in claim 6, wherein the calculation unit comprises:

a first retrieval sub-unit that, according to the preset diversity field, classifies the first search result set to obtain a respective subset of the plurality of subsets corresponding to each respective attribute category of the first search result set;

a second retrieval sub-unit that, according to the first relevance score in each subset, obtains a ranking position of the respective first search result; and

a matching unit that, according to a preset relationship between the ranking position of the respective first search result and the second relevance score, obtains the second relevance score of the respective first search result.

8. The system as recited in claim 6, further comprising:

a store unit that stores the query data, the preset number of second search results, and a corresponding relationship between the query data and the preset number of second search results into a database.

9. The system as recited in claim 6, wherein the retrieval unit comprises:

a searching sub-unit that, according to the first relevance score, conducts a search based on the query data submitted by the client; and

a second extraction sub-unit that, according to the preset diversity field, extracts first search results from search results of the search.

10. The system as recited in claim 6, further comprising:

a display unit that displays the preset number of second search results to the client.

11. One or more storage devices storing executable instructions that, when executed by a computer, instruct the computer to perform acts comprising:

according to query data submitted from a client, obtaining a first search result set of first search results relevant to the query data, the first search results included in the first search result set being ranked according to first relevance scores of the first search results;

calculating second relevance scores of the first search results based on a function of ranking positions of the first search results within respective subsets of the first search result set that are classified according to the preset diversity fields of the first search results;

according to the first relevance scores and the second relevance score of the first search results, generating relevance parameter scores for the first search results, wherein the generating comprises summing a respective first relevance score and a respective second relevance score to provide a respective relevance parameter score for each first search result; and

according to the relevance parameter scores of the first search results, extracting a preset number of second search results from the first search result set for presentation to the client, wherein the extracting comprises;

according to the relevance parameter score, ranking each subset of the respective subsets after classification of the first search results; and

according to a ranking order, extracting the preset number of second search results respectively from the ranked subsets, the preset number of second search results being a product of a number of diversity values and a number of recurring extractions.

12. The one or more storage devices as recited in claim 11, wherein calculating the second relevance score of each first search result comprises:

according to the preset diversity fields of the first search results in the first search result set, classifying the first search result set to obtain the respective subsets corresponding to attribute categories of the first search result set;

according to the first relevance scores, obtaining the ranking positions of the first search results within the respective subsets of the first search result set; and

according to a preset relationship between a ranking position and a second relevance score, obtaining the second relevance scores of the first search results.

13. The one or more storage devices as recited in claim 11, further comprising:

storing the query data, the preset number of second search results, and a corresponding relationship between the query data and the preset number of second search results into a database.

14. The one or more storage devices as recited in claim 11, wherein obtaining the first search result set of first search results relevant to the query data comprises:

according to the first relevance scores, conducting a search based on the query data submitted from the client; and

according to the preset diversity fields, extracting the first search results from search results of the search.