# US 2019/0377740 A1 · Measuring the influence of entities over an audience on a topic

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

- Number: US 2019/0377740 A1
- Document ID: US20190377740A1
- Status: Published application
- Kind: A1
- Application number: 16/547,753
- Filed: 2019-08-22
- Published: 2019-12-12
- Earliest claimed: 2015-09-29
- Assignees: International Business Machines Corporation
- Applicants: International Business Machines Corporation
- Inventors: Alfredo Alba; Clemens Drews; Daniel Gruhl; Neal R. Lewis; Pablo N. Mendes; Meenakshi Nagarajan; CARTIC RAMAKRISHNAN
- Web page: https://patentgrep.com/patent/US20190377740A1

## CPC

- G06Q 30/0201: Market modelling; Market analysis; Collecting market data
- G06F 16/243: {Natural language query formulation}
- G06F 16/24573: {using data annotations, e.g. user-defined metadata}
- G06F 16/24575: {using context}
- G06F 16/24578: {using ranking}
- G06F 16/285: {Clustering or classification}
- G06F 16/951: Indexing; Web crawling techniques
- G06Q 10/46: Determination of level of influence of users within social networking services
- G06Q 10/48: using social graphs
- G06Q: INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)

## Abstract

Influencers (individuals or groups) over a selected audience (observers or recipients of information, objects and/or events) on a given topic are measured based on influence features, which include a sentiment flipping influence feature indicative of ability of an audience member to influence other audience members to change their sentiment on the selected topic. Other influence features include the ability to influence others: to change followership; to express interest in a topic associated with a hashtag pioneered by the influencer, based on the effectiveness and phrasing of language used. The output of the influence engine can be a score representing the relative influence of audience members over the audience on the topic of interest. Influencers may be ranked according to their total influence score over the audience on the topic.

## Claims

### Claim 1

A computer implemented method comprising:
receiving an influencer request comprising a selected audience, a selected topic, and a selected date range;
simultaneously calculating component influence feature scores for each audience member of the selected audience corresponding to the selected topic and the date range, and storing the calculated component influence feature scores in accessible computer memory, wherein calculating the component influence feature scores comprises:
simultaneously determining a sentiment flipping influence feature score for each audience member of the selected audience corresponding to the selected topic and the selected date range;
simultaneously determining a hashtag pioneer influence feature score for each audience member of the selected audience corresponding to the selected topic and the selected date range;
simultaneously determining a followership conversion influence feature score for each audience member of the selected audience corresponding to the selected topic and the selected date range;
simultaneously determining a conversation influence feature score for each audience member of the selected audience corresponding to the selected topic and the selected date range;
simultaneously determining an edge and node statistics influence feature score for each audience member of the selected audience corresponding to the selected topic and the selected date range;
simultaneously determining an effectiveness and phrasing influence feature score for each audience member of the selected audience corresponding to the selected topic and the selected date range;
simultaneously determining a total influence score for each audience member of the selected audience corresponding to the selected topic and the received date range based on all of the component influence feature scores stored in the computer memory; and
in accordance with the influencer request, identifying one or more audience members as the influencers of the selected audience on the selected topic within the selected date range, based on the total influence scores of each audience member.

### Claim 2 (depends on 1)

The computer implemented method of claim 1, wherein the sentiment flipping influence feature indicates an extent to which an audience member has influenced at least one other audience member to change sentiment on the selected topic.

### Claim 3 (depends on 1)

The computer implemented method of claim 1, wherein the hashtag pioneer influence feature indicates an extent to which at least one audience member expresses an interest in the selected topic associated with a hashtag created by each audience member.

### Claim 4 (depends on 1)

The computer implemented method of claim 1, wherein the followership conversion influence feature indicates an extent to which each audience member influenced the at least one other audience member to change followership.

### Claim 5 (depends on 1)

The computer implemented method of claim 1, wherein the conversation influence feature indicates an extent that a conversation on the selected topic involving one audience member influences one or more audience members to engage in that conversation.

### Claim 6 (depends on 1)

The computer implemented method of claim 1, wherein the effectiveness and phrasing influence feature indicates an extent to which at least one audience member was influenced by the effectiveness and phrasing of language used by each audience member.

### Claim 7 (depends on 1)

The computer implemented method of claim 1, wherein the edge and node statistics influence feature indicates an extent that audience members are influenced to forward or reply to a message on the topic originated by one audience member.

### Claim 8 (depends on 1)

The computer implemented method of claim 1, wherein calculating the component influence feature scores comprises: determining an adoption influence feature score for each audience member of the selected audience corresponding to the selected topic, wherein the adoption influence feature indicates an extent one audience member influences other audience members to adopt a certain position on the selected topic.

### Claim 9 (depends on 1)

The computer implemented method of claim 1, wherein calculating the component influence feature scores comprises: determining a network attributes influence feature score for each audience member of the selected audience corresponding to the selected topic, wherein the network attributes influence feature indicates an extent of homophily between audience members on the selected topic.

### Claim 10 (depends on 1)

The computer implemented method of claim 1, further comprising:
comparing one of the component influence scores for at least one audience member with a first predetermined threshold value, and increasing the one component influence score for the at least one audience member in response to the one component influence score being above the first predetermined threshold value; and
comparing one of the component influence scores for at least one audience member with a second predetermined threshold value, and decreasing the one component influence score for the at least one audience member in response to the one component influence score being below the second predetermined threshold value.

### Claim 11

A computer implemented method comprising:
receiving an influencer request identifying an audience, a topic, and a date range;
simultaneously calculating a component influence feature score for each audience member of the audience corresponding to the topic and date range, wherein calculating the component influence feature scores comprises:
simultaneously determining a sentiment flipping influence feature score for each audience member corresponding to the topic and the date range, and storing the sentiment flipping influence feature scores in accessible computer memory;
simultaneously determining a hashtag pioneer influence feature score for each audience member corresponding to the topic and the date range, and storing the hashtag pioneer influence feature scores in accessible computer memory;
simultaneously determining a followership conversion influence feature score for each audience member corresponding to the topic and the date range, and storing the followership conversion influence feature scores in accessible computer memory;
simultaneously determining an effectiveness and phrasing influence feature score for each audience member corresponding to the topic and the date range, and storing the effectiveness and phrasing influence feature scores in accessible computer memory;
comparing one of the component influence scores for at least one audience member with a first predetermined threshold value, and increasing the one component influence score for the at least one audience member in response to the one component influence score being above the first predetermined threshold value;
comparing one of the component influence scores for at least one audience member with a second predetermined threshold value, and decreasing the one component influence score for the at least one audience member in response to the one component influence score being below the second predetermined threshold value;
determining a total influence score for each audience member corresponding to the topic and the date range based on all of the component influence feature scores stored in the computer memory; and
in accordance with the influencer request, identifying one or more audience members as the influencers of the selected audience on the selected topic within the selected date range, based on the total influence scores of each audience member.

### Claim 12 (depends on 11)

The computer implemented method of claim 11, wherein calculating the component influence feature scores comprises: determining a conversation influence feature score for each audience member of the audience corresponding to the topic and the date range, wherein the conversation influence feature indicates an extent that a conversation on the topic involving one audience member influences one or more audience members to engage in that conversation.

### Claim 13 (depends on 11)

The computer implemented method of claim 11, wherein calculating the component influence feature scores comprises: determining an edge and node statistics influence feature score for each audience member of the audience corresponding to the topic and the date range, wherein the edge and node statistics influence feature indicates an extent that audience members are influenced to forward or reply to a message on the topic originated by one audience member.

### Claim 14 (depends on 11)

The computer implemented method of claim 11, wherein calculating the component influence feature scores comprises: determining an adoption influence feature score for each audience member of the audience corresponding to the topic and the date range, wherein the adoption influence feature indicates an extent one audience member influences other audience members to adopt a certain position on the selected topic.

### Claim 15 (depends on 11)

The computer implemented method of claim 11, wherein calculating the component influence feature scores comprises: determining a network attributes influence feature score for each audience member of the audience corresponding to the topic and the date range, wherein the network attributes influence feature indicates an extent of homophily between audience members on the topic.

### Claim 16 (depends on 11)

The computer implemented method of claim 11, wherein the sentiment flipping influence feature indicates an extent to which an audience member has influenced at least one other audience member to change sentiment on the topic.

### Claim 17 (depends on 11)

The computer implemented method of claim 11, wherein the hashtag pioneer influence feature indicates an extent to which at least one audience member expresses an interest in the topic associated with a hashtag created by each audience member.

### Claim 18 (depends on 11)

The computer implemented method of claim 11, wherein the followership conversion influence feature indicates an extent to which each audience member influenced the at least one other audience member to change followership.

### Claim 19 (depends on 11)

The computer implemented method of claim 11, wherein the effectiveness and phrasing influence feature indicates an extent to which at least one audience member was influenced by the effectiveness and phrasing of language used by each audience member.