# US 2017/0255653 A1 · Method for categorizing images to be associated with content items based on keywords of search queries

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

- Number: US 2017/0255653 A1
- Document ID: US20170255653A1
- Status: Published application
- Kind: A1
- Application number: 15/057,968
- Filed: 2016-03-01
- Published: 2017-09-07
- Earliest claimed: 2016-03-01
- Assignees: Baidu USA LLC
- Applicants: Baidu USA LLC
- Inventors: Fan Zhu; Qi Kong; Li Zhuang; Xiang Yu; Lei Ma; Dong Li; Yifei Jiang; Shuang Wu; Guang Yang; Jingao Wang
- Web page: https://patentgrep.com/patent/US20170255653A1

## CPC

- G06F 16/55: Clustering; Classification
- G06F 16/51: Indexing; Data structures therefor; Storage structures
- G06F 16/5866: {using information manually generated, e.g. tags, keywords, comments, manually generated location and time information}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

According to one embodiment, a first keyword-to-image (keyword/image) mapping table is provided. The first keyword mapping table includes a number of entries, each entry mapping a keyword to one or more image identifiers (IDs) identifying one or more images. For each of the keywords of the first keyword/image mapping table, an analysis is performed on the keyword to determine one or more related keywords that are related to the keyword. One or more additional entries corresponding to the one or more related keywords are generated to be incorporated into the first keyword/image mapping table to generate a second keyword/image mapping table. The second keyword/image mapping table is utilized to associated a particular image to a particular content item related to a particular keyword.

## Claims

### Claim 1

A computer-implemented method for categorizing images to be associated with content items based on keywords, the method comprising:
providing a first keyword-to-image (keyword/image) mapping table, the first keyword mapping table including a plurality of entries, each entry mapping a keyword to one or more image identifiers (IDs) identifying one or more images;
for each of the keywords of the first keyword/image mapping table, performing, by an analysis module executed by a processor, an analysis on the keyword to determine one or more related keywords that are related to the keyword; and
generating, by a configuration module executed by the processor, one or more additional entries corresponding to the one or more related keywords to be incorporated into the first keyword/image mapping table to generate a second keyword/image mapping table, wherein the second keyword/image mapping table is utilized to associated a particular image to a particular content item related to a particular keyword.

### Claim 2 (depends on 1)

The method of claim 1, wherein each of the entries of the first keyword/image mapping table maps a single keyword to one or more image IDs.

### Claim 3 (depends on 1)

The method of claim 1, wherein at least one of the entries of the second keyword/image mapping table maps a phrase having a plurality of keywords to one or more image IDs.

### Claim 4 (depends on 1)

The method of claim 1, wherein the one or more additional entries map the related keywords to identical image IDs of the keyword of the first keyword/image mapping table.

### Claim 5 (depends on 1)

The method of claim 1, wherein performing an analysis on the keyword to determine one or more related keywords comprises performing a semantic analysis on each keyword of the first keyword/image mapping table to derive one or more related keywords that are semantically similar to the keyword.

### Claim 6 (depends on 1)

The method of claim 1, further comprising:
for each of the additional entries in the second keyword/image mapping table, performing, by a filtering module executed by the processor, a second analysis on the related keywords in view of image metadata of corresponding images to verify whether the related keywords describe the corresponding images;
removing at least one image from the additional entries whose related keywords do not sufficiently describe the at least one image based on the second analysis; and
generating a third keyword/image mapping table, wherein the third keyword/image mapping table is utilized to associated a particular image to a particular content item related to a particular keyword.

### Claim 7 (depends on 6)

The method of claim 6, wherein performing a second analysis comprising:
examining a source from which a first image in a first entry of the second keyword/image mapping table was obtained;
determining whether the source of the first image contains a word that is related to at least one of the keywords in the first entry; and
removing the first image from the first entry of the second keyword/image mapping table in response to determining that the source does not contain any keyword of the first entry.

### Claim 8 (depends on 6)

The method of claim 6, wherein performing a second analysis comprising:
examining a source from which a first image in a first entry of the second keyword/image mapping table was obtained;
determining an access frequency representing how frequent the first image has been accessed by users at the source; and
removing the first image of the first entry if the access frequency is below a predetermine threshold.

### Claim 9

A non-transitory machine-readable medium having instructions stored therein, which when executed by a processor, cause the processor to perform operations of categorizing images to be associated with content items based on keywords, the operations comprising:
providing a first keyword-to-image (keyword/image) mapping table, the first keyword mapping table including a plurality of entries, each entry mapping a keyword to one or more image identifiers (IDs) identifying one or more images;
for each of the keywords of the first keyword/image mapping table, performing an analysis on the keyword to determine one or more related keywords that are related to the keyword; and
generating one or more additional entries corresponding to the one or more related keywords to be incorporated into the first keyword/image mapping table to generate a second keyword/image mapping table, wherein the second keyword/image mapping table is utilized to associated a particular image to a particular content item related to a particular keyword.

### Claim 10 (depends on 9)

The non-transitory machine-readable medium of claim 9, wherein each of the entries of the first keyword/image mapping table maps a single keyword to one or more image IDs.

### Claim 11 (depends on 9)

The non-transitory machine-readable medium of claim 9, wherein at least one of the entries of the second keyword/image mapping table maps a phrase having a plurality of keywords to one or more image IDs.

### Claim 12 (depends on 9)

The non-transitory machine-readable medium of claim 9, wherein the one or more additional entries map the related keywords to identical image IDs of the keyword of the first keyword/image mapping table.

### Claim 13 (depends on 9)

The non-transitory machine-readable medium of claim 9, wherein performing an analysis on the keyword to determine one or more related keywords comprises performing a semantic analysis on each keyword of the first keyword/image mapping table to derive one or more related keywords that are semantically similar to the keyword.

### Claim 14 (depends on 9)

The non-transitory machine-readable medium of claim 9, wherein the operations further comprise:
for each of the additional entries in the second keyword/image mapping table, performing a second analysis on the related keywords in view of image metadata of corresponding images to verify whether the related keywords describe the corresponding images;
removing at least one image from the additional entries whose related keywords do not sufficiently describe the at least one image based on the second analysis; and
generating a third keyword/image mapping table, wherein the third keyword/image mapping table is utilized to associated a particular image to a particular content item related to a particular keyword.

### Claim 15 (depends on 14)

The non-transitory machine-readable medium of claim 14, wherein performing a second analysis comprising:
examining a source from which a first image in a first entry of the second keyword/image mapping table was obtained;
determining whether the source of the first image contains a word that is related to at least one of the keywords in the first entry; and
removing the first image from the first entry of the second keyword/image mapping table in response to determining that the source does not contain any keyword of the first entry.

### Claim 16 (depends on 14)

The non-transitory machine-readable medium of claim 14, wherein performing a second analysis comprising:
examining a source from which a first image in a first entry of the second keyword/image mapping table was obtained;
determining an access frequency representing how frequent the first image has been accessed by users at the source; and
removing the first image from the first entry if the access frequency is below a predetermine threshold.

### Claim 17

A data processing system, comprising:
a processor; and
a memory coupled to the processor to store a first keyword-to-image (keyword/image) mapping table, the first keyword mapping table including a plurality of entries, each entry mapping a keyword to one or more image identifiers (IDs) identifying one or more images;
an analysis module, for each of the keywords of the first keyword/image mapping table, to perform an analysis on the keyword to determine one or more related keywords that are related to the keyword; and
a configuration module to generate one or more additional entries corresponding to the one or more related keywords to be incorporated into the first keyword/image mapping table to generate a second keyword/image mapping table, wherein the second keyword/image mapping table is utilized to associated a particular image to a particular content item related to a particular keyword.

### Claim 18 (depends on 17)

The system of claim 17, wherein each of the entries of the first keyword/image mapping table maps a single keyword to one or more image IDs.

### Claim 19 (depends on 17)

The system of claim 17, wherein at least one of the entries of the second keyword/image mapping table maps a phrase having a plurality of keywords to one or more image IDs.

### Claim 20 (depends on 17)

The system of claim 17, wherein the one or more additional entries map the related keywords to identical image IDs of the keyword of the first keyword/image mapping table.

### Claim 21 (depends on 17)

The system of claim 17, wherein the analysis module is to perform a semantic analysis on each keyword of the first keyword/image mapping table to derive one or more related keywords that are semantically similar to the keyword.

### Claim 22 (depends on 17)

The system of claim 17, further comprising a filtering module to:
for each of the additional entries in the second keyword/image mapping table, perform a second analysis on the related keywords in view of image metadata of corresponding images to verify whether the related keywords describe the corresponding images,
remove at least one image from the additional entries whose related keywords do not sufficiently describe the at least one image based on the second analysis, and
generate a third keyword/image mapping table, wherein the third keyword/image mapping table is utilized to associated a particular image to a particular content item related to a particular keyword.

### Claim 23 (depends on 22)

The system of claim 22, wherein in performing a second analysis, the filtering module is further to:
examine a source from which a first image in a first entry of the second keyword/image mapping table was obtained,
determine whether the source of the first image contains a word that is related to at least one of the keywords in the first entry, and
remove the first image from the first entry of the second keyword/image mapping table in response to determining that the source does not contain any keyword of the first entry.

### Claim 24 (depends on 22)

The system of claim 22, wherein in performing a second analysis, the filtering module is further to:
examine a source from which a first image in a first entry of the second keyword/image mapping table was obtained,
determine an access frequency representing how frequent the first image has been accessed by users at the source, and
remove the first image from the first entry if the access frequency is below a predetermine threshold.
