Searching based on an identifier of a searcher

• US 9,465,875 B2
• Filed: 05/31/2012
• Issued: 10/11/2016
• Est. Priority Date: 05/31/2012
• Status: Active Grant

• Current Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Original Assignee: Hewlett Packard Enterprise Development LP (Hewlett-Packard Enterprise Company)
• Inventors: Barkol, Omer, Golan, Shahar, Aharon, Michal, Vexler, Reuth
• Primary Examiner(s): ASPINWALL, EVAN S
• Application Number:
  US13/484,396
• Publication Number:
  US 20130325852A1
• Time in Patent Office: 1,594 Days
• Field of Search: 707/723
• US Class Current: 1/1
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation

Abstract

A query is received to search data, where the query includes a search term. A search of the data is performed in response to the query, wherein the search produces result data based on the search term and an identifier of a searcher submitting the query.

Claims

1. A method of multi-phased personalized searching, comprising:

performing by a computer system;

performing a search of data in response to a first query and an identifier of a searcher, wherein the search produces result data based on a search term in the first query and the identifier of the searcher;

selecting a first subset of terms in data items of a subset of data items in the result data;

computing correlation weights for corresponding terms in the first subset of terms, where each of the correlation weights is based on correlation of the corresponding term to data items in the subset of data items;

selecting a second subset of the terms in the first subset of terms according to the correlation weights;

producing a second query containing the second subset of terms; and

processing the second query to perform further searching that retrieves data items matching the second subset of terms from public websites available over an Internet,

wherein the multi-phase personalized searching includes;

a first search phase comprising receiving the first query, obtaining the identifier of the searcher, and the performing of the search of the data in response to the first query and the identifier of the searcher, and

a second search phase comprising the selecting of the first subset of terms, the computing of the correlation weights, the selecting of the second subset of the terms in the first subset of terms, and the producing of the second query.

2. The method of claim 1, wherein at least a portion of the result data includes content involving the searcher identified by the identifier, and wherein the content involving the searcher includes at least one selected from among a data item authored by the searcher, a data item received by the searcher, and a data item produced based on participation of the searcher.

3. The method of claim 1, wherein receiving the first query includes receiving the first query responsive to the search input including a search string entered by the searcher.

4. The method of claim 1, wherein receiving the first query includes receiving the first query that contains terms of an input data item for use in identifying at least one result data item that is similar to the input data item.

5. The method of claim 1, further comprising:

ranking data items in the result data according to relevancy of the data items to the search performed in response to the first query; and

selecting the subset of the data items in the result data based on the ranking.

6. The method of claim 1, wherein the identifier of the searcher is obtained independently of a search input provided by the searcher to produce the first query.

7. An article comprising at least one non-transitory machine-readable storage medium storing instructions to perform multi-phase personalized searching, the instructions upon execution by a computer system causing the computer system to:

perform a first search of a corpus of data based on a first query containing a search term, and based on an identifier of a searcher that submitted the first query;

rank data items in result data produced by the first search, the ranking according to relevance of the data items to the first search;

select a subset of the data items in the result data based on the ranking;

select a first subset of terms in data items of the subset of data items;

compute correlation weights for corresponding terms in the first subset of terms, where each of the correlation weights is based on correlation of the corresponding term to data items in the subset of data items;

select a second subset of the terms in the first subset of terms according to the correlation weights;

produce a second query that contains the terms in the second subset of terms; and

perform a second search of the corpus of data based on the second query, the second search retrieving data items matching the second subset of terms in the second query from public websites available over an Internet.

8. The article of claim 7, wherein the terms in the subset of data items include at least one term in a data item received by the searcher.

9. The article of claim 7, wherein the first query contains the search term entered as a search string by the searcher.

10. The article of claim 7, wherein the first query contains search terms extracted from an input data item.

11. The article of claim 10, wherein the second search produces further result data containing data items that are similar to the input data item.

12. The article of claim 7, wherein the first search produces the result data from public websites available over the Internet based on the search term and the identifier of the searcher.

13. A system comprising:

at least one processor to perform multi-phase personalized searching comprising;

a first search phase comprising;

receiving a first query to search data, the first query including terms of an input data item, where the first query is to identify result data items that are similar to the input data item, and

performing a search of the data in response to the first query, wherein the search produces result data based on the terms and an identifier of a searcher submitting the first query, where at least a portion of the result data includes content involving the searcher identified by the identifier; and

a second search phase comprising;

selecting a first subset of terms in data items of a subset of data items in the result data,

computing correlation weights for corresponding terms in the first subset of terms, where each of the correlation weights is based on correlation of the corresponding term to data items in the subset of data items,

selecting a second subset of the terms in the first subset of terms according to the correlation weights,

creating a second query containing the second subset of terms; and

returning further result data produced by a second search responsive to the second query, the second search performed after the search responsive to the first query.

14. The system of claim 13, wherein the second search retrieves data items matching the second subset of terms in the second query from public websites available over an Internet.

15. The system of claim 13, wherein the identifier of the searcher is obtained independently of a search input provided by the searcher to produce the first query.

16. The system of claim 13, wherein the search in response to the first query is performed using the terms of the first query and the identifier of the searcher against data sources including public websites available over an Internet.

17. The system of claim 13, wherein the at least one processor is to further:

rank data items in the result data produced by the search in response to the first query, the ranking according to relevancy of the data items to the search performed in response to the first query;

select the subset of the data items in the result data based on the ranking.