# US 2010/0169300 A1 · Ranking Oriented Query Clustering and Applications

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

- Number: US 2010/0169300 A1
- Document ID: US20100169300A1
- Status: Published application
- Kind: A1
- Application number: 12/345,645
- Filed: 2008-12-29
- Published: 2010-07-01
- Earliest claimed: 2008-12-29
- Assignees: Microsoft Corporation
- Inventors: Qi Liu; Ruihua Song; Jiangming Yang
- Web page: https://patentgrep.com/patent/US20100169300A1

## CPC

- G06F 16/9566: {URL specific, e.g. using aliases, detecting broken or misspelled links}
- G06F 16/951: Indexing; Web crawling techniques
- G06F 16/9532: Query formulation
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Techniques described herein allow for suggesting creation of tools for improving search engine performance. Specifically, these tools focus on producing more relevant search engine results via a URL-based query clustering method. These tools first extract tokens from Uniform Resource Locators associated to search queries. With these tokens, these tools form query clusters of common tokens. The resulting clusters can be used to help understand the similarities in user search queries via URL-based cluster queries to produce more relevant search results.

## Claims

### Claim 1

A method of associating a search query to a cluster comprising:
receiving a Uniform Resource Locator (URL) returned in response to a search based on a search query;
extracting a first set of tokens from the URL, wherein the extracting comprises:
i. splitting the URL by predefined separators;
ii. removing URL stop-words; and
iii. merging identical tokens;
determining a similarity between the first set of tokens and each of multiple set of tokens extracted from multiple other URLs returned in response to other search queries; and
associating the search query with a cluster, wherein the associating is based at least in part on the similarities between the first set of tokens and each of the multiple set of tokens extracted from the other URLs.

### Claim 2 (depends on 1)

The method of claim 1, wherein the determining the similarity between the first set of tokens and each of a multiple set of tokens utilizes a vector space model.

### Claim 3 (depends on 1)

The method of claim 1, wherein the determining the similarity between the first set of tokens and each of a multiple set of tokens utilizes a set model.

### Claim 4 (depends on 1)

The method of the claim 1, further comprising assigning the cluster to a cluster category.

### Claim 5 (depends on 1)

The method of claim 1, wherein the tokens are words extracted from a URL.

### Claim 6

A method of associating a search query to a cluster comprising:
receiving a Uniform Resource Locator (URL) returned in response to a search based on a search query;
extracting a first set of tokens from the URL;
determining a similarity between the first set of tokens and each of a multiple set of tokens extracted from multiple other URLs returned in response to other search queries; and
associating the search query with a cluster, wherein the associating is based at least in part on the similarities between the first set of tokens and each of the multiple set of tokens extracted from the other URLs.

### Claim 7 (depends on 6)

The method of the claim 6, wherein the extracting comprises:
splitting the URL by predefined separators; and
removing numerical characters.

### Claim 8 (depends on 6)

The method of the claim 6, wherein the extracting further comprises removing URL stop-words.

### Claim 9 (depends on 6)

The method of the claim 6, wherein the extracting further comprises merging identical tokens.

### Claim 10 (depends on 6)

The method of claim 6, wherein the determining the similarity between the first set of tokens and each of a multiple set of tokens utilizes a set model.

### Claim 11 (depends on 6)

The method of the claim 6, further comprising assigning the cluster to a cluster category.

### Claim 12 (depends on 6)

The method of claim 6, wherein the tokens are words extracted from a URL.

### Claim 13 (depends on 6)

The method of the claim 6, further comprising determining relevant websites for a specific cluster.

### Claim 14

One or more computer-readable media storing computer-executable instructions that, when executed on one or more processors, perform acts comprising:
extracting a first set of tokens from a Uniform Resource Locator (URL), wherein the extracting comprises:
i. splitting the URL by predefined separators; and
ii. removing numerical characters;
determining a similarity between the first set of tokens and each of a multiple set of tokens extracted from multiple other URLs; and
associating the search query with a cluster, wherein the associating is based at least in part on the similarities between the first set of tokens and each of the multiple set of tokens extracted from the other URLs.

### Claim 15 (depends on 14)

One or more computer-readable media as recited in claim 14, wherein
the extracting further comprises removing URL stop-words.

### Claim 16 (depends on 14)

One or more computer-readable media as recited in claim 14, wherein the extracting further comprises merging identical tokens.

### Claim 17 (depends on 14)

One or more computer-readable media as recited in claim 14, wherein the determining the similarity between the first set of tokens and each of a multiple set of tokens utilizes a set model.

### Claim 18 (depends on 14)

One or more computer-readable media as recited in claim 14, further comprising assigning the cluster to a cluster category.

### Claim 19 (depends on 14)

One or more computer-readable media as recited in claim 14, wherein the tokens are words extracted from a URL.

### Claim 20 (depends on 14)

One or more computing devices comprising:
one or more processors; and
the one or more computer-readable media storing the computer-executable instructions as recited in claim 14.
