Method and System of Information Matching in Electronic Commerce Website

• US 20120143816A1
• Filed: 08/18/2010
• Published: 06/07/2012
• Est. Priority Date: 08/27/2009
• Status: Active Grant

• Current Assignee: Alibaba Group Holding Ltd.
• Original Assignee: Alibaba Group Holding Ltd.
• Inventors: Zhang, Xu, Liu, Qing-Yan, Wu, Peng-Song, Ye, Yi-Huo
• Granted Patent: US 8,346,782 B2
• Time in Patent Office: Days
• US Class Current: 707/607
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9538 Presentation of query results
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising

Abstract

The present disclosure describes an information matching method and information system. The method collects characteristic data of category plurality of categories of network behavior of a plurality of network users; groups the network users into a plurality of clusters based on the characteristic data; sets weight of the characteristic data; receives a search request from one of the plurality of network users; obtains search results according to the search request; inquires a historical record of clicking operation on at least some of the search results made by at least some of the network users in one of the clusters which contains the one of the plurality of network users; calculates a respective rank value of each of the at least some of the search results based on the historical record of clicking operation of at least some of the network users and the weight of the characteristic data; sorts of the search results to arrange the search results according to the rank values of the search results in a descending order from high rank to low rank; and causes the sorted search results to be displayed.

Claims

1. An information matching method, comprising:

collecting characteristic data of a plurality of categories of network behavior of a plurality of network users;

grouping the network users into a plurality of clusters based on the characteristic data;

setting weight of the characteristic data;

receiving a search request from a specific network user of the plurality of network users;

obtaining search results according to the search request;

inquiring a historical record of clicking operation on at least some of the search results made by at least some of the network users in one of the clusters which contains the specific network user;

calculating a respective rank value of the at least some of the search results based on the historical record of clicking operation of the at least some of the network users and the weight of the characteristic data;

sorting the at least some of the search results to arrange the at least some of the search results according to the respective rank values in a descending order from high rank to low rank; and

causing the sorted search results to be displayed.

2. The method of claim 1, wherein the network behavior includes online trading or web comments, and wherein the characteristic data includes online transaction record or web comments record.

3. The method of claim 1, wherein grouping the network users into a plurality of clusters based on the characteristic data comprises:

grouping those network users of the plurality of network users having no characteristic data of network behavior collected into a first cluster;

grouping at least some of the remaining network users according to the characteristic data of network behavior into a pre-determined sets of clusters; and

storing results of clustering in a database as a data sheet.

4. The method of claim 3, wherein grouping the network users into a plurality of clusters based on the characteristic data comprises:

if the characteristic data of the network behavior is online transaction record, grouping the network users based on whether information of commodity in the online transaction record is similar to at least some other network users and the network users who have purchased similar commodity are grouped into a same cluster; and

completing clustering when a number of clusters attained reaches a pre-determined number.

5. The method of claim 3, wherein grouping the network users into a plurality of clusters based on the characteristic data comprises:

if the characteristic data of the network behavior is web comments record, grouping the network users based on the cluster of a merchant user to which a web comment is made by at least some of the network user;

or

if the characteristic data of the network behavior is web comments record, grouping the network users by making statistics of a number of network users of the plurality of network users whose comments appear in the web comments record of every two merchant users to obtain overlap proportion based on the number of network users and a total number of network users who made web comments against the merchant user;

calculating a space between merchant users;

carrying out clustering based on the space and then carrying out clustering based on the clustering of merchant users.

6. The method of claim 1, wherein collecting characteristic data of category plurality of categories of network behavior of a plurality of network users comprises:

collecting the characteristic data through server log analysis system, network activities log system, geographical information system, interface of a third party, or a combination thereof.

7. The method of claim 1, further comprising:

setting a weight of geographical position information;

calculating the rank value of at least some of the search results based on the weight of geographical position information and the weight of the characteristic data; and

sorting the search results based on the calculated rank values to arrange the search results in a descending order.

8. The method of claim 1, wherein receiving a search request from one of the plurality of network users comprises receiving a key word or clicking by the one of the plurality of network users as the search request.

9. An information matching system for electronic commerce website, comprising:

an information acquisition system that collects characteristic data of a plurality of categories of network behavior of a plurality of network users, carries out clustering of the network users according to the characteristic data and against the plurality of categories of network behavior, and sets weight of the characteristics data for carrying out clustering;

a retrieval system that receives a search request of a specific network user of the plurality of network users, obtains search results, and inquires a historical record of clicking operation on at least some of the search results made by other network users in a cluster containing the specific network user, calculates a respective rank value of each of the at least some of the search results based on the historical record of clicking operation and the weight of the characteristic data for carrying out the clustering, and sorts the search results of according to the calculated rank values to arrange the search results in a descending order from high rank to low rank;

a results page generating system that displays the sorted search results.

10. The system of claim 9, wherein the retrieval system comprises:

a search engine server that receives the search request of the specific network user, and obtains the search results according to the search request;

a sorting system that inquires the historical record of clicking operation of at least some of the search results made by other network users in the cluster containing the specific network user, calculates the rank values of the search results according to the historical record of clicking operation made by the other network users and based on the weight of the characteristic data for carrying out clustering, and sorts the search results according to the rank values to arrange the search results in the descending order.

11. The system of claim 10, wherein the sorting system comprises:

a first setting module that sets the weight of the characteristic data for carrying out clustering;

a inquiring module that inquires the historical record of clicking operation made by at least some of the network users against at least one of the search results;

a statistic module that performs statistic analysis of the historical record of clicking operation of at least some of the search results to provide statistic data, and keeps the statistic data in a database in a form of data sheet; and

a sorting module that inquires, against the search results, the historical record of clicking operation made by at least some of the network users in the cluster containing the specific network user, calculates the respective rank value of each of the at least some of the search results according to the weight of the characteristic data, and sorts the search results based on the calculated rank values to arrange the search results in the descending order.

12. The system of claim 11, wherein the sorting system further comprises:

a second setting mould that sets a weight of geographical position information, wherein;

the sorting module inquires, against the search results, the historical record of clicking operation made by at least some of the network users in the cluster containing the specific network user, calculates the respective rank value of each of the at least some of the search results based on the weight of geographical position information and the weight of characteristic data for carrying out clustering, and sorts the search results to arrange the search results in the descending order.