Relativistic sentiment analyzer

Patent metadata

Field	Value
Publication number	US-10546235-B2
Application number	US-201615098201-A
Country	US
Kind code	B2
Filing date	Apr 13, 2016
Priority date	Apr 8, 2015
Publication date	Jan 28, 2020
Grant date	Jan 28, 2020

Abstract

Sentiment analyzer systems may include feedback analytics servers configured to receive and analyze feedback data from various client devices. Feedback data may be received and analyzed to determine feedback context and sentiment scores. In some embodiments, natural language processing neural networks may be used to determine sentiment scores for the feedback data. Feedback data also may be grouped into feedback aggregations based on context, and sentiment scores may be calculated for each feedback aggregation. Sentiment analyzer outputs and corresponding output devices may be determined based on the sentiment scores and feedback contexts.

First claim

What is claimed is:

1. A method of analyzing feedback data, comprising:
determining, based on feedback data received from one or more client devices, one or more sentiment scores associated with one or more users, relating to a content distribution network;
receiving user records relating to one or more interactions of the one or more users with the content distribution network, the interactions occurring after the receipt of the feedback data;
training a machine learning algorithm based on the received user records and the sentiment scores associated with the one or more users;
receiving subsequent feedback data associated with a first user, relating to the content distribution network;
calculating a sentiment score for the first user, based on the subsequent feedback data associated with the first user;
determining a user record prediction for the first user, based on the sentiment score calculated for the first user, using the trained machine learning algorithm; and
determining and providing an output to one or more output devices, based on the determined user record prediction for the first user.

2. The method of claim 1 , wherein determining the one or more sentiment scores based on the feedback data comprises:
providing the feedback data to a natural language processing (NLP) neural network; and
receiving the one or more sentiment scores from the NLP neural network corresponding to the feedback data.

3. The method of claim 1 , wherein the feedback data received from the one or more client devices comprises multimodal user input data relating to the content distribution network, and wherein determining the one or more sentiment scores based on the feedback data comprises at least two of:
using a language processing engine to determine a sentiment score for text feedback data;
using a voice analyzer to determine a sentiment score for voice feedback data;
using a gesture analyzer to determine a sentiment score for movement feedback data; and
using an eye movement analyzer to determine a sentiment score for eye movement feedback data.

4. The method of claim 1 , further comprising:
grouping the feedback data received from the one or more client devices into a plurality feedback aggregations, said grouping comprising at least one of:
grouping the feedback data by user, into at least a first feedback aggregation associated with the first user and a second feedback aggregation associated with a second user;
grouping the feedback data by content item, into at least a first aggregation of feedback from a plurality of users associated with a first content item of the content distribution network and a second aggregation of feedback from the plurality of users associated with a second content item of the content distribution network; or
grouping the feedback data by time, into at least a first feedback aggregation associated with the first user during a first time period and a second feedback aggregation associated with the first user during a second time period.

5. The method of claim 1 , further comprising:
grouping the feedback data received from the one or more client devices into at least a first feedback aggregation associated with the first user during a first time period, and a second feedback aggregation associated with the first user during a second time period;
calculating separate sentiment scores for the first user for the first time period and the second time period, using the first feedback aggregation and the second feedback aggregation;
determining a change in sentiment score for the first user the first time period and the second feedback time period; and
in response to determining that the change in sentiment score for the first user the first time period and the second feedback time period exceeds a threshold, initiating an intervention process for the first user.

6. The method of claim 1 , further comprising:
grouping the feedback data received from the one or more client devices into at least a first feedback aggregation corresponding to feedback from a plurality of users associated with a first content item of the content distribution network, and a second feedback aggregation corresponding to feedback from the plurality of users associated with a second content item of the content distribution network;
calculating separate sentiment scores for the first feedback aggregation associated with the first content item, and the second feedback aggregation associated with the second content item;
determining a difference in sentiment scores for the plurality of users associated with the first content item and the second content item; and
providing output including the difference in sentiment scores for the plurality of users associated with the first content item and the second content item.

7. The method of claim 1 , wherein the feedback data received from the one or more client devices is related to a presentation of live content via the content distribution network, and wherein determining and providing the output to the one or more output devices comprises:
determining one or more computing devices associated with a presenter of the live content via the content distribution network;
determining the output for the presenter of the live content, based on at least one of the calculated sentiment score for the first user and the determined user record prediction for the first user; and
transmitting the output to the computing devices associated with the presenter of the live content, during the presentation of the live content via the content distribution network.

8. The method of claim 1 , further comprising:
receiving the feedback from the one or more client devices via an event streaming service executing on a data store server.

9. A feedback analytics server for a content distribution network, comprising:
a processing unit comprising one or more processors; and
memory coupled with and readable by the processing unit and storing therein a set of instructions which, when executed by the processing unit, causes the feedback analytics server to:
determine, based on feedback data received from one or more client devices, one or more sentiment scores associated with one or more users, relating to the content distribution network;
receive user records relating to one or more interactions of the one or more users with the content distribution network, the interactions occurring after the receipt of the feedback data;
train a machine learning algorithm based on the received user records and the sentiment scores associated with the one or more users;
receive subsequent feedback data associated with a first user, relating to the content distribution network;
calculate a sentiment score for the first user, based on the subsequent feedback data associated with the first user;
determine a user record prediction for the first user, based on the sentiment score calculated for the first user, using the trained machine learning algorithm; and
determine and provide an output to one or more output devices, based on the determined user record prediction for the first user.

10. The feedback analytics server of claim 9 , wherein determining the one or more sentiment scores based on the feedback data comprises:
providing the feedback data to a natural language processing (NLP) neural network; and
receiving the one or more sentiment scores from the NLP neural network corresponding to the feedback data.

11. The feedback analytics server of claim 9 , wherein the feedback data received from the one or more client devices comprises multimodal user input data relating to the content distribution network, and wherein determining the one or more sentiment s

Assignees

• Pearson Education Inc

Inventors

• Moudy Christopher
• Paterson Todd
• Berns Kevin

Classifications

• G06F40/30

  Semantic analysis · CPC title

• G06F40/40

  Processing or translation of natural language (natural language analysis G06F40/20; semantic analysis G06F40/30) · CPC title

• G06F16/24578

  using ranking · CPC title

• G06N3/084

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

• G06F16/244

  Grouping and aggregation · CPC title