# US 2018/0121429 A1 · Returning search results utilizing topical user click data when search queries are dissimilar

> Agents: this data is available as JSON from the patentgrep API and MCP server — see https://patentgrep.com/llms.txt

- Number: US 2018/0121429 A1
- Document ID: US20180121429A1
- Status: Published application
- Kind: A1
- Application number: 15/336,107
- Filed: 2016-10-27
- Published: 2018-05-03
- Earliest claimed: 2016-10-27
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: Priyanka Agrawal; Man Chu; Steven M. Pritko; Justin A. Ziniel
- Web page: https://patentgrep.com/patent/US20180121429A1

## CPC

- G06F 16/3326: {using relevance feedback from the user, e.g. relevance feedback on documents, documents sets, document terms or passages}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

A data handling system includes at least two host devices. For a particular search query that is received by a first host device, a preliminary set of search results is therein generated. The first host system maps the query to one or more topics that are representative to the query. The first host system provides topical click data associated with the topic to a machine-learning module located within a second host device that determines a relevancy score of the result utilizing the topical click data. The first host system obtains the relevancy score of the result and re-ranks the order of the result within a set of results based upon the relevancy score.

## Claims

### Claim 1

A method of re-ranking search results by utilizing topical click data in an environment where search result queries are ostensibly dissimilar, the method comprising:
determining, with a first host device, a preliminary result [R] of a search query [Q];
determining, with the first host device, a topic [T] that is representative of the search query [Q];
retrieving, with the first host device, topical click data [C] associated with the topic [T] and associated with the result [R];
determining, with a second host device connected to the first host device by a network, a relevance value [V] of the result [R] utilizing the topical click data [C]; and
re-ranking, with the first host device, the order of the preliminary result [R] within a set of results based upon the relevance value [V].

### Claim 2 (depends on 1)

The method of claim 1, wherein retrieving topical click data [C] associated with the topic [T] comprises:
reading a topical data structure from a storage device of the first host device, wherein the topical data structure comprises a plurality of topical click data associated with the topic [T] and associated with the result [R].

### Claim 3 (depends on 2)

The method of claim 2, wherein the storage device of the first host device comprises a plurality of topical data structures, wherein each topical data structure is associated with a unique topic [T] and comprises a plurality of topical click data associated with the unique topic [T] and various results.

### Claim 4 (depends on 1)

The method of claim 1, further comprising:
subsequent to determining the preliminary result [R] of the search query [Q], saving a <Q,R> data pair to a storage device of the first host device;
subsequent to determining the topic [T] that is representative of the search query [Q], saving a <Q,R,T> data triplet to the storage device of the first host device;
subsequent to retrieving topical click data [C] associated with the topic [T], saving a <Q,R,T,C> data quadruple to the storage device of the first host device; and
subsequent to receiving the relevance value [V] from the second host device, saving a <Q,R,T,C,V> data quintuple to the storage device of the first host device.

### Claim 5 (depends on 1)

The method of claim 1, wherein determining the relevance value [V] of the result [R] utilizing the topical click data [C] comprises:
compiling, with the second host device, a test search query result comprising a plurality of results;
sending, with the second host device, the test search query result to a tester;
receiving, with the second host device, a re-ranked test search query result comprising a re-ranked plurality of results from the tester;
determining, with the second host device, logic utilized by the tester to re-rank the plurality of results;
emulating, with the second host device, the determined logic utilized by the tester to re-rank the plurality of results to determine the relevance value [V] of the result [R].

### Claim 6 (depends on 1)

The method of claim 1, wherein the topic [T] is a single word topic that is most representative of the search query [Q].

### Claim 7 (depends on 1)

The method of claim 1, wherein the topic [T] is a plurality of single word topics [T1, Tn] that are representative of the search query [Q].

### Claim 8 (depends on 7)

The method of claim 7, wherein the single word topic [T1] is weighted as more representative of the search query [Q] relative to the single word topic [Tn].

### Claim 9 (depends on 8)

The method of claim 8, further comprising:
attempting to retrieve, with the first host device, topical click data [C1] associated with the single topic [T1];
determining, with the first host device, that no topical click data [C1] is associated with the single topic [T1]; and
subsequent to determining that no topical click data [C1] is associated with the single topic [T1], retrieving, with host device, topical click data [Cn] associated with the single topic [Tn].

### Claim 10 (depends on 1)

The method of claim 1, wherein the first host device receives a plurality of queries that each include a plurality of words organized in a character string and wherein the plurality of queries share less than a quarter percent of characters.

### Claim 11

A computer program product for re-ranking search results by utilizing topical click data in an environment where search result queries are ostensibly dissimilar, the computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions readable to cause:
a first host device to determine a preliminary result [R] of a search query [Q];
the first host device to determine a topic [T] that is representative of the search query [Q];
the first host device to retrieve topical click data [C] associated with the topic [T] and associated with the result [R];
a second host device to determine a relevance value [V] of the result [R] utilizing the topical click data [C], wherein the first host device and the second host device are connected by a network; and
the first host device to re-rank the order of the preliminary result [R] within a set of results based upon the relevance value [V].

### Claim 12 (depends on 11)

The computer program product of claim 11, wherein the program instructions that cause the first host device to retrieve topical click data [C] associated with the topic [T] further cause the host device to read a topical data structure from a storage device of the first host device, wherein the topical data structure comprises a plurality of topical click data associated with a topic [T] and associated with the result [R].

### Claim 13 (depends on 12)

The computer program product of claim 12, wherein the storage device of the first host device comprises a plurality of topical data structures, wherein each topical data structure is associated with a unique topic [T] and wherein each topical data structure comprises a plurality of topical click data associated with the unique topic [T] and associated with various results.

### Claim 14 (depends on 11)

The computer program product of claim 11, wherein the program instructions are readable to cause:
subsequent to determining the preliminary result [R] of the search query [Q], the first host device to locally save a <Q,R> data pair;
subsequent to determining the topic [T] that is representative of the search query [Q], the first host device to locally save a <Q,R,T> data triplet;
subsequent to retrieving topical click data [C] associated with the topic [T], the first host device to locally save a <Q,R,T,C> data quadruple; and
subsequent to receiving the relevance value [V] of the result [R] from the second host device, the first device to locally save a <Q,R,T,C,V> data quintuple to the storage device.

### Claim 15 (depends on 11)

The computer program product of claim 11, wherein the program instructions that cause the second host device to determine the relevance value [V] of the result [R] utilizing the topical click data [C] further cause the second host device to:
compile a test search query result comprising a plurality of results;
send the test search query result to a tester;
receive a re-ranked test search query result comprising a re-ranked plurality of results from the tester;
determine logic utilized by the tester to re-rank the plurality of results;
emulate the determined logic utilized by the tester to re-rank the plurality of results to determine the relevance value [V] of the result [R].

### Claim 16 (depends on 11)

The computer program product of claim 11, wherein the topic [T] is a single word topic that is most representative of the search query [Q].

### Claim 17 (depends on 11)

The computer program product of claim 11, wherein the topic [T] is a plurality of single word topics [T1, Tn] that are representative of the search query [Q].

### Claim 18 (depends on 17)

The computer program product of claim 17, wherein the single word topic [T1] is weighted as more representative of the search query [Q] relative to the single word topic [Tn].

### Claim 19 (depends on 17)

The computer program product of claim 17, wherein the program instructions are readable by the first host device to further cause the first host device to:
attempt to retrieve topical click data [C1] associated with the single word topic [T1];
determine that no topical click data [C1] is associated with the single topic [T1]; and
subsequent to determining that no topical click data [C1] is associated with the single topic [T1], retrieve topical click data [Cn] associated with the single topic [Tn].

### Claim 20

A data handling system comprising:
a first host device comprising a first processor and first memory communicatively coupled to the first processor; and
a second host device comprising a second processor and second memory communicatively coupled to the second processor,
wherein the first memory is encoded with instructions, wherein the first instructions are executable by the first processor to cause the first processor to:
determine a preliminary result [R] of a search query [Q];
determine a topic [T] that is representative of the search query [Q];
retrieve topical click data [C] associated with the topic [T] and associated with the result [R];
wherein the second memory is encoded with second instructions, wherein the second instructions are executable by the second processor to cause the second processor to:
determine a relevance value [V] of the result [R] utilizing the topical click data [C];
and wherein the first instructions are executable by the first processor to further cause the first processor to:
re-rank the order of the preliminary result [R] within a set of results based upon the relevance value [V].