# US 2018/0322212 A1 · Network Resource Recommendation Method And Computer Device

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

- Number: US 2018/0322212 A1
- Document ID: US20180322212A1
- Status: Published application
- Kind: A1
- Application number: 16/038,892
- Filed: 2018-07-18
- Published: 2018-11-08
- Earliest claimed: 2016-01-19
- Assignees: Huawei Technologies Co., Ltd.
- Applicants: Huawei Technologies Co., Ltd.
- Inventors: Guilin LI
- Web page: https://patentgrep.com/patent/US20180322212A1

## CPC

- G06F 16/9535: Search customisation based on user profiles and personalisation
- G06F 16/00: Information retrieval; Database structures therefor; File system structures therefor
- G06F 16/24578: {using ranking}
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

The present invention discloses a network resource recommendation method, including: determining a first interest model according to a user behavior record in a current time window; calculating a similarity between any two interest models in the first interest model, a second interest model, and a third interest model, where a time window corresponding to the second interest model is adjacent to the current time window, the third interest model is determined based on all user behavior records in a reference time window set, the reference time window set includes N time windows prior to the time window corresponding to the second interest model; determining a fourth interest model and a recommendation algorithm according to the calculated similarity between any two interest models; generating a network resource recommendation list according to the fourth interest model and the recommendation algorithm; and making a recommendation according to the network resource recommendation list.

## Claims

### Claim 1

A network resource recommendation method, comprising:
determining a first interest model according to a user behavior record in a current time window;
calculating a similarity between any two interest models in the first interest model, a second interest model, and a third interest model, wherein a time window corresponding to the second interest model is adjacent to the current time window, the third interest model is an interest model determined based on all user behavior records in a reference time window set, the reference time window set comprises N time windows prior to the time window corresponding to the second interest model, and N is an integer that is greater than 1;
determining a fourth interest model and a recommendation algorithm according to the calculated similarity between any two interest models;
generating a network resource recommendation list according to the fourth interest model and the recommendation algorithm; and
making a recommendation according to the network resource recommendation list.

### Claim 2 (depends on 1)

The method according to claim 1, wherein the calculated similarity between any two interest models comprises a first similarity, a second similarity, and a third similarity, the first similarity is a similarity between the first interest model and the second interest model, the second similarity is a similarity between the first interest model and the third interest model, and the third similarity is a similarity between the second interest model and the third interest model.

### Claim 3 (depends on 2)

The method according to claim 2, wherein the determining a fourth interest model according to the calculated similarity between any two interest models comprises:
if both the first similarity and the second similarity are greater than or equal to a preset similarity, determining the fourth interest model according to the second interest model, the third interest model, and a first formula;
if the first similarity is greater than or equal to the preset similarity, and the second similarity is less than the preset similarity, determining the fourth interest model according to the first interest model, the second interest model, and the first formula;
if the first similarity is less than the preset similarity, and the second similarity is greater than or equal to the preset similarity, determining the fourth interest model according to the first interest model, the third interest model, and the first formula;
if the first similarity is less than the preset similarity, the second similarity is less than the preset similarity, and the third similarity is greater than or equal to the preset similarity, determining the fourth interest model according to the second interest model, the third interest model, and the first formula; or
if the first similarity is less than the preset similarity, the second similarity is less than the preset similarity, and the third similarity is less than the preset similarity, using the third interest model as the fourth interest model, wherein
the first formula is:
M=αA+βB
wherein M indicates the fourth interest model, A and B are any two interest models in the first interest model, the second interest model, and the third interest model, α indicates a weight of A, β indicates a weight of B, and a sum of α and β is equal to 1.

### Claim 4 (depends on 3)

The method according to claim 3, wherein after the determining the fourth interest model according to the second interest model, the third interest model, and a first formula, the method further comprises:
updating the third interest model to the fourth interest model.

### Claim 5 (depends on 3)

The method according to claim 3, wherein after the determining the fourth interest model according to the first interest model, the second interest model, and the first formula, the method further comprises:
determining a fifth interest model according to the second interest model, the third interest model, and the first formula, and updating the third interest model to the fifth interest model.

### Claim 6 (depends on 2)

The method according to claim 2, wherein the determining a recommendation algorithm according to the calculated similarity between any two interest models comprises:
using a content-based collaborative filtering recommendation as the recommendation algorithm if both the first similarity and the second similarity are greater than or equal to a preset similarity;
using a content-based recommendation as the recommendation algorithm if the first similarity is greater than or equal to the preset similarity, and the second similarity is less than the preset similarity; or
using a collaborative filtering-based recommendation as the recommendation algorithm if the first similarity is less than the preset similarity, and the second similarity is greater than or equal to the preset similarity, or if the first similarity is less than the preset similarity, and the second similarity is less than the preset similarity, and the third similarity is greater than or equal to the preset similarity, or if the first similarity is less than the preset similarity, and the second similarity is less than the preset similarity, and the third similarity is less than the preset similarity.

### Claim 7 (depends on 1)

The method according to claim 1, wherein the interest model may be a vector model, and the calculating a similarity between any two interest models in the first interest model, a second interest model, and a third interest model comprises:
calculating, according to a second formula, a similarity between any two vector models in a first vector model, a second vector model, and a third vector model, wherein
the second formula is:
similarity
=
A
·
B
||
A
||
·
||
B
||
=
∑
i
=
1
n


A
i
×
B
i
∑
i
=
1
n


(
A
i
)
2
×
∑
i
=
1
n


(
B
i
)
2
wherein the similarity indicates the similarity between any two vector models, A and B are any two vector models in the first vector model, the second vector model, and the third vector model, n indicates a total quantity of elements comprised in a vector model, and i indicates an ith element in the vector model.

### Claim 8

A computer device, comprising:
a processor and a memory, wherein the processor invokes code or an instruction in the memory to perform the following steps:
determining a first interest model according to a user behavior record in a current time window;
calculating a similarity between any two interest models in the first interest model, a second interest model, and a third interest model, wherein a time window corresponding to the second interest model is adjacent to the current time window, the third interest model is an interest model determined based on all user behavior records in a reference time window set, the reference time window set comprises N time windows prior to the time window corresponding to the second interest model, and N is an integer that is greater than 1;
determining a fourth interest model and a recommendation algorithm according to the calculated similarity between any two interest models;
generating a network resource recommendation list according to the fourth interest model and the recommendation algorithm; and
making a recommendation according to the network resource recommendation list.

### Claim 9 (depends on 8)

The computer device according to claim 8, wherein the calculated similarity between any two interest models comprises a first similarity, a second similarity, and a third similarity, the first similarity is a similarity between the first interest model and the second interest model, the second similarity is a similarity between the first interest model and the third interest model, and the third similarity is a similarity between the second interest model and the third interest model.

### Claim 10 (depends on 9)

The computer device according to claim 9, wherein a specific implementation in which the processor determines the fourth interest model according to the calculated similarity between any two interest models is:
if both the first similarity and the second similarity are greater than or equal to a preset similarity, determining the fourth interest model according to the second interest model, the third interest model, and a first formula;
if the first similarity is greater than or equal to the preset similarity, and the second similarity is less than the preset similarity, determining the fourth interest model according to the first interest model, the second interest model, and the first formula;
if the first similarity is less than the preset similarity, and the second similarity is greater than or equal to the preset similarity, determining the fourth interest model according to the first interest model, the third interest model, and the first formula;
if the first similarity is less than the preset similarity, the second similarity is less than the preset similarity, and the third similarity is greater than or equal to the preset similarity, determining the fourth interest model according to the second interest model, the third interest model, and the first formula; or
if the first similarity is less than the preset similarity, the second similarity is less than the preset similarity, and the third similarity is less than the preset similarity, using the third interest model as the fourth interest model, wherein
the first formula is:
M=αA+βB
wherein M indicates the fourth interest model, A and B are any two interest models in the first interest model, the second interest model, and the third interest model, α indicates a weight of A, β indicates a weight of B, and a sum of α and β is equal to 1.

### Claim 11 (depends on 10)

The computer device according to claim 10, wherein after the processor determines the fourth interest model according to the second interest model, the third interest model, and the first formula, the processor is further configured to:
update the third interest model to the fourth interest model.

### Claim 12 (depends on 10)

The computer device according to claim 10, wherein after the processor determines the fourth interest model according to the first interest model, the second interest model, and the first formula, the processor is further configured to:
determine a fifth interest model according to the second interest model, the third interest model, and the first formula, and update the third interest model to the fifth interest model.

### Claim 13 (depends on 9)

The computer device according to claim 9, wherein a specific implementation in which the processor determines the recommendation algorithm according to the calculated similarity between any two interest models is:
using a content-based collaborative filtering recommendation as the recommendation algorithm if both the first similarity and the second similarity are greater than or equal to a preset similarity;
using a content-based recommendation as the recommendation algorithm if the first similarity is greater than or equal to the preset similarity, and the second similarity is less than the preset similarity; or
using a collaborative filtering-based recommendation as the recommendation algorithm if the first similarity is less than the preset similarity, and the second similarity is greater than or equal to the preset similarity, or if the first similarity is less than the preset similarity, and the second similarity is less than the preset similarity, and the third similarity is greater than or equal to the preset similarity, or if the first similarity is less than the preset similarity, and the second similarity is less than the preset similarity, and the third similarity is less than the preset similarity.

### Claim 14 (depends on 8)

The computer device according to claim 8, wherein the interest model may be a vector model, and a specific implementation in which the processor calculates the similarity between any two interest models in the first interest model, the second interest model, and the third interest model is:
calculating, according to a second formula, a similarity between any two vector models in a first vector model, a second vector model, and a third vector model, wherein
the second formula is:
similarity
=
A
·
B
||
A
||
·
||
B
||
=
∑
i
=
1
n


A
i
×
B
i
∑
i
=
1
n


(
A
i
)
2
×
∑
i
=
1
n


(
B
i
)
2
wherein the similarity indicates the similarity between any two vector models, A and B are any two vector models in the first vector model, the second vector model, and the third vector model, n indicates a total quantity of elements comprised in a vector model, and i indicates an ith element in the vector model.
