# US 2017/0286535 A1 · Systems and methods for identifying issues in electronic documents

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

- Number: US 2017/0286535 A1
- Document ID: US20170286535A1
- Status: Published application
- Kind: A1
- Application number: 15/629,321
- Filed: 2017-06-21
- Published: 2017-10-05
- Earliest claimed: 2013-09-05
- Assignees: TSG Technologies, LLC
- Applicants: TSG Technologies, LLC
- Inventors: Noah Krusell; Anthony L. Hinrichs; Nathan D. Dees; Caimei Lu
- Web page: https://patentgrep.com/patent/US20170286535A1

## CPC

- G06F 16/35: Clustering; Classification
- G06Q 10/10: Office automation; Time management
- G06F: ELECTRIC DIGITAL DATA PROCESSING (computer systems based on specific computational models G06N)
- 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

## Abstract

A computer system identifies and prioritizes issues in electronic documents from electronic social media sources and/or from traditional media sources. The computer system includes a communication network linked to one or more of the sources and a computer server in communication with the communication network. The computer server is configured to receive electronic documents from the sources via the communication network. The computer server has memory and a processor, the memory including processor executable instructions stored in the memory and executable by the processor. The instructions comprise data collection instructions, topic modeling instructions including instructions for weighing and instructions for indicating a distribution for identified issues within each of the retrieved documents, issue assignment instructions, instructions for determining, and instructions for ranking.

## Claims

### Claim 1

A computer system for identifying and prioritizing issues in electronic documents from electronic social media sources and/or from traditional media sources, the computer system comprising:
a communication network linked to one or more of the sources;
a computer server in communication with the communication network, the computer server to receive electronic documents from the sources via the communication network;
the computer server having memory and a processor, the memory including processor executable instructions stored in the memory and executable by the processor, the instructions comprising:
data collection instructions for retrieving and scoring documents from the sources via the network;
topic modeling instructions for identifying one or more issues in each of the retrieved documents according to terms in the retrieved documents including:
instructions for weighing each term in each of the retrieved documents with regard to relevance of the term to each identified issue;
instructions for indicating a distribution for identified issues within each of the retrieved documents;
issue assignment instructions for formally associating a limited list of specific issues with each retrieved document;
instructions for determining summary metrics for each identified issue; and
instructions for ranking the issues relative to each other based on weighted summary metrics.

### Claim 2 (depends on 1)

The system of claim 1 wherein the topic modeling instructions comprise:
initial topic modeling instructions for identifying one or more initial issues in the collection of the retrieved documents according to terms in the retrieved documents; and
subtopic modeling instructions for identifying one or more issues in each of the initial issues of the retrieved documents according to terms in the retrieved documents for each initial issue.

### Claim 3 (depends on 1)

The system of claim 1 wherein the instructions for weighing each term in each of the retrieved documents with regard to the relevance of the term to each identified issue comprises instructions for determining a topic-term probability matrix, and wherein the instructions for indicating a distribution for identified issues within each of the retrieved documents comprises instructions for determining a document-topic probability matrix.

### Claim 4 (depends on 1)

The system of claim 1 wherein the issue assignment instructions comprise instructions for limiting the number of issues formally associated with each of the retrieved documents based on the values in the document-topic probability matrix.

### Claim 5 (depends on 1)

The system of claim 1 wherein instructions for the determining summary metrics comprise:
core message instructions for determining a Volume V and a Tone T for each issue based on the number of retrieved documents, and terms in the retrieved documents, respectively
influencer instructions for determining an Influencer network strength I for each issue based on the entities identified within the retrieved documents;
speed instructions for determining a Speed S for each issue based on a timing of the terms in the retrieved documents; and
emotion instructions for determining an Emotion E for each issue.

### Claim 6 (depends on 1)

The system of claim 1 wherein the instructions for ranking comprise instructions for determining a weighted sum of the Volume V, the Tone T, the Influencer network strength I, the Speed S, and the Emotion E of each identified issue.

### Claim 7 (depends on 6)

The system of claim 6 wherein the instructions for determining a weighted sum comprise:
Instructions for determining a single quantitative ranking score, R, for each issue (i) according to the following:
R
i
=
ω
V

(
V
i
-
μ

(
V
)
σ

(
V
)
)
-
ω
T

(
T
i
-
μ

(
T
)
σ

(
T
)
)
+
ω
I

(
I
i
-
μ

(
I
)
σ

(
I
)
)
+
ω
S

(
S
i
-
μ

(
S
)
σ

(
S
)
)
+
ω
E

(
E
i
-
μ

(
E
)
σ

(
E
)
)
(
Equation


1
)
wherein each summary metric is weighted using weights ωv, ωT, ωs, ωE and ωI for Volume, Tone, Influencer network strength, Speed, and Emotion, respectively, and wherein each summary metric is scaled by subtraction of a mean (μ) across all of the issues and division by a standard deviation (σ) across all of the issues.

### Claim 8 (depends on 7)

The system of claim 7 further comprising instructions for selecting the weights ωV, ωT, ωI, ωS and ωE based upon an iterative process of evaluating historical quantitative scores calculated for previously-discovered issues, and comparing the scores for accuracy to the actual lifecycles of the described issues.

### Claim 9

A non-transitory computer readable storage medium for use with a computer system for identifying and prioritizing issues in electronic documents from electronic social media sources and/or from traditional media sources, wherein the computer system includes a communication network linked to one or more of the sources, and a computer server in communication with the communication network, the computer server to receive electronic documents from the sources via the communication network, the computer server having a processor;
the non-transitory computer readable storage medium including processor executable instructions stored on the non-transitory computer readable storage medium and executable by the processor, the instructions comprising:
data collection instructions for retrieving and scoring documents from the sources via the network;
topic modeling instructions for identifying one or more issues in each of the retrieved documents according to terms in the retrieved documents including:
instructions for weighing each term in each of the retrieved documents with regard to relevance of the term to each identified issue;
instructions for indicating a distribution for identified issues within each of the retrieved documents;
issue assignment instructions for formally associating a limited list of specific issues with each retrieved document;
instructions for determining summary metrics for each identified issue; and
instructions for ranking the issues relative to each other based on weighted summary metrics.

### Claim 10 (depends on 9)

The medium of claim 9 wherein the topic modeling instructions comprise:
initial topic modeling instructions for identifying one or more initial issues in the collection of the retrieved documents according to terms in the retrieved documents; and
subtopic modeling instructions for identifying one or more issues in each of the initial issues of the retrieved documents according to terms in the retrieved documents for each initial issue.

### Claim 11 (depends on 9)

The medium of claim 9 wherein the instructions for weighing each term in each of the retrieved documents with regard to the relevance of the term to each identified issue comprises instructions for determining a topic-term probability matrix, and wherein the instructions for indicating a distribution for identified issues within each of the retrieved documents comprises instructions for determining a document-topic probability matrix.

### Claim 12 (depends on 9)

The medium of claim 9 wherein the issue assignment instructions comprise instructions for limiting the number of issues formally associated with each of the retrieved documents based on the values in the document-topic probability matrix.

### Claim 13 (depends on 9)

The medium of claim 9 wherein instructions for the determining summary metrics comprise:
core message instructions for determining a Volume V and a Tone T for each issue based on the number of retrieved documents, and terms in the retrieved documents, respectively;
influencer instructions for determining an Influencer network strength I for each issue based on the entities identified within the retrieved documents;
speed instructions for determining a Speed S for each issue based on a timing of the terms in the retrieved documents; and
emotion instructions for determining an Emotion E for each issue.

### Claim 14 (depends on 9)

The medium of claim 9 wherein the instructions for ranking comprise instructions for determining a weighted sum of the Volume V, the Tone T, the Influencer network strength I, the Speed S, and the Emotion E of each identified issue.

### Claim 15 (depends on 14)

The medium of claim 14 wherein the instructions for determining a weighted sum comprise:
instructions for determining a single quantitative ranking score, R, for each issue (i) according to the following:
R
i
=
ω
V

(
V
i
-
μ

(
V
)
σ

(
V
)
)
-
ω
T

(
T
i
-
μ

(
T
)
σ

(
T
)
)
+
ω
I

(
I
i
-
μ

(
I
)
σ

(
I
)
)
+
ω
S

(
S
i
-
μ

(
S
)
σ

(
S
)
)
+
ω
E

(
E
i
-
μ

(
E
)
σ

(
E
)
)
(
Equation


1
)
wherein each summary metric is weighted using weights ωV, ωT, ωS, ωE and ωI for Volume, Tone, Influencer network strength, Speed, and Emotion, respectively, and wherein each summary metric is scaled by subtraction of a mean (μ) across all of the issues and division by a standard deviation (σ) across all of the issues.

### Claim 16 (depends on 15)

The medium of claim 15 further comprising instructions for selecting the weights ωV, ωT, ωI, ωS and ωE based upon an iterative process of evaluating historical quantitative scores calculated for previously-discovered issues, and comparing the scores for accuracy to the actual lifecycles of the described issues.

### Claim 17

A computer executable storage method for identifying and prioritizing issues in electronic documents from electronic social media sources and/or from traditional media sources, wherein a communication network is linked to one or more of the sources, the method including processor executable instructions stored on a non-transitory computer readable storage medium and executable by a processor, the method comprising:
retrieving and scoring documents from the sources via the network;
identifying one or more issues in each of the retrieved documents according to terms in the retrieved documents including:
weighing each term in each of the retrieved documents with regard to a relevance of the term to each identified issue;
indicating a distribution for identified issues within each of the retrieved documents;
formally associating a limited list of specific issues with each retrieved document;
determining summary metrics for each identified issue; and
ranking the issues relative to each other based on weighted summary metrics.

### Claim 18 (depends on 17)

The method of claim 17 wherein the determining summary metrics comprises:
determining a Volume V and a Tone T for each issue based on the number of retrieved documents, and terms in the retrieved documents, respectively
determining an Influencer network strength I for each issue based on the entities identified within the retrieved documents;
determining a Speed S for each issue based on a timing of the terms in the retrieved documents; and
determining an Emotion E for each issue.

### Claim 19 (depends on 17)

The method of claim 17 wherein the ranking comprises determining a weighted sum of the Volume V, the Tone T, the Influencer network strength I, the Speed S, and the Emotion E of each identified issue.

### Claim 20 (depends on 19)

The method of claim 19 wherein the determining a weighted sum comprises:
determining a single quantitative ranking score, R, for each issue (i) according to the following:
R
i
=
ω
V

(
V
i
-
μ

(
V
)
σ

(
V
)
)
-
ω
T

(
T
i
-
μ

(
T
)
σ

(
T
)
)
+
ω
I

(
I
i
-
μ

(
I
)
σ

(
I
)
)
+
ω
S

(
S
i
-
μ

(
S
)
σ

(
S
)
)
+
ω
E

(
E
i
-
μ

(
E
)
σ

(
E
)
)
(
Equation


1
)
wherein each summary metric is weighted using weights ωV, ωT, ωS, ωE and ωI for Volume, Tone, Influencer network strength, Speed, and Emotion, respectively, and wherein each summary metric is scaled by subtraction of a mean (μ) across all of the issues and division by a standard deviation (σ) across all of the issues.