Predicting alternative communication based on textual analysis

Patent metadata

Field	Value
Publication number	US-12131259-B2
Application number	US-202017109034-A
Country	US
Kind code	B2
Filing date	Dec 1, 2020
Priority date	Dec 1, 2020
Publication date	Oct 29, 2024
Grant date	Oct 29, 2024

Abstract

A processor-implemented method for predicting alternative communications based on textual analysis. The method includes building, by machine learning, a model to predict an optimal communication method, whereby the building includes training the model on a knowledge corpus of historic data and user data, and results of previous predictions in similar circumstances. The method further includes intercepting textual communication within communication channels, wherein the intercepting comprises a keyboard capture, a screen capture, or both a keyboard capture and a screen capture. The method further includes identifying, by pattern analysis, sentiment analysis, and textual analysis, topics, sentiments, and participants within the intercepted textual communication. The method further includes predicting, by the model, the optimal communication method, whereby the optimal communication method comprises continuing the textual communication, a video conference or a telephone conference.

First claim

What is claimed is:

1. A processor-implemented method for predicting alternative communications based on textual analysis, the method comprising:
building, by machine learning, a model to predict an optimal communication method, wherein the building includes training the model on a knowledge corpus of historic data and user data, and results of previous predictions in similar circumstances;
intercepting textual communication in real time between a plurality of participants within communication channels, wherein the intercepting is selected from the group consisting of a keyboard capture, a screen capture, or both, the keyboard capture and the screen capture;
identifying, by pattern analysis, sentiment analysis, and textual analysis, topics, sentiments, and the plurality of participants within the intercepted textual communication;
predicting, by the model, the optimal communication method; and
generating an overlay notification for the plurality of participants, based on the predicted optimal communication method being other than continuing the textual communication, wherein the generated overlay notification is configured to enable the plurality of participants to accept or reject the predicted optimal communication method.

2. The method of claim 1 , further comprising:
based on the plurality of participants accepting the predicted optimal communication method, generating calendar entries for each participant; and
based on the plurality of participants rejecting the predicted optimal communication method, continuing the textual communication.

3. The method of claim 1 , wherein the knowledge corpus is stored in a cloud computing segment by topic, wherein the topic is determined by topical analysis of the knowledge corpus, and wherein a pointer is stored in the knowledge corpus to point to the corresponding cloud computing segment storing the topic.

4. The method of claim 1 , wherein the knowledge corpus initially includes enterprise repositories, wherein the enterprise repositories comprise email archives, technical problem reports, and employee online directories, and wherein the knowledge corpus is updated through a feedback loop with data describing factors identified in previous interactions between different participants, an amount of time for each interaction, and whether the optimal communication method was accepted or rejected.

5. The method of claim 4 , wherein the factors identified in the previous interactions are selected from the group consisting of topic, area of expertise of the plurality of participants, length of time to resolve past similar issues, method of communication, and whether the resolution was successful.

6. The method of claim 4 , wherein the factors are assigned weights that are fine-tuned by a back-propagation technique based on error rates obtained in previous runs of the method.

7. The method of claim 1 , wherein a time saved measurement is determined by averaging past recorded times, wherein the past recorded times are selected from the group consisting of user time to read email, user time to send emails, and total elapsed time between conversations.

8. A computer system for predicting alternative communications for time saving in issue resolution, the computer system comprising:
one or more processors, one or more computer-readable memories, one or more computer-readable tangible storage medium, and program instructions stored on at least one of the one or more tangible storage medium for execution by at least one of the one or more processors via at least one of the one or more memories, wherein the computer system is capable of performing a method comprising:
building, by machine learning, a model to predict an optimal communication method, wherein the building includes training the model on a knowledge corpus of historic data and user data, and results of previous predictions in similar circumstances;
intercepting textual communication in real time between a plurality of participants within communication channels, wherein the intercepting is selected from the group consisting of a keyboard capture, a screen capture, or both, the keyboard capture and the screen capture;
identifying, by pattern analysis, sentiment analysis, and textual analysis, topics, sentiments, and the plurality of participants within the intercepted textual communication;
predicting, by the model, the optimal communication method; and
generating an overlay notification for the plurality of participants, based on the predicted optimal communication method being other than continuing the textual communication, wherein the generated overlay notification is configured to enable the plurality of participants to accept or reject the predicted optimal communication method.

9. The computer system of claim 8 , further comprising:
based on the plurality of participants accepting the predicted optimal communication method, generating calendar entries for each participant; and
based on the plurality of participants rejecting the predicted optimal communication method, continuing the textual communication.

10. The computer system of claim 8 , wherein the knowledge corpus is stored in a cloud computing segment by topic, wherein the topic is determined by topical analysis of the knowledge corpus, and wherein a pointer is stored in the knowledge corpus to point to the corresponding cloud computing segment storing the topic.

11. The computer system of claim 8 , wherein the knowledge corpus initially includes enterprise repositories, wherein the enterprise repositories comprise email archives, technical problem reports, and employee online directories, and wherein the knowledge corpus is updated through a feedback loop with data describing factors identified in previous interactions between different participants, an amount of time for each interaction, and whether the optimal communication method was accepted or rejected.

12. The computer system of claim 11 , wherein the factors identified in the previous interactions are selected from the group consisting of topic, area of expertise of the plurality of participants, length of time to resolve past similar issues, method of communication, and whether the resolution was successful.

13. The computer system of claim 11 , wherein the factors are assigned weights that are fine-tuned by a back-propagation technique based on error rates obtained in previous runs of the method.

14. The computer system of claim 8 , wherein a time saved measurement is determined by averaging past recorded times, wherein the past recorded times are selected from the group consisting of user time to read email, user time to send emails, and total elapsed time between conversations.

15. A computer program product for predicting alternative communications for time saving in issue resolution, the computer program product comprising:
one or more computer-readable tangible storage medium and program instructions stored on at least one of the one or more tangible storage medium, the program instructions executable by a processor, the program instructions comprising:
building, by machine learning, a model to predict an optimal communication method, wherein the building includes training the model on a knowledge corpus of historic data and user data, and results of previous predictions in similar circumstances;
intercepting textual communication in real time between a plurality of participants within communication channels, wherein the intercepting is selected from the group consisting of a keyboard capture, a screen capture, or both, the keyboard capture and the screen capture;
identifying, by pattern a

Assignees

• IBM

Inventors

• Decrop Clement
• Agrawal Tushar
• Fox Jeremy R
• Rakshit Sarbajit K

Classifications

• G06N3/0442

  characterised by memory or gating, e.g. long short-term memory [LSTM] or gated recurrent units [GRU] · CPC title

• G06N3/09

  Supervised learning · CPC title

• G06N3/044

  Recurrent networks, e.g. Hopfield networks · CPC title

• G06F40/30 Primary

  Semantic analysis · CPC title

• G06N3/084 Primary

  Backpropagation, e.g. using gradient descent · CPC title