Predicting human behavior by machine learning of natural language interpretations

• US 10,296,586 B2
• Filed: 02/06/2017
• Issued: 05/21/2019
• Est. Priority Date: 12/23/2016
• Status: Active Grant

• Current Assignee: Soundhound AI IP LLC
• Original Assignee: SoundHound Incorporated
• Inventors: Singh, Pranav, Mont-Reynaud, Bernard, Khov, Kheng, Probell, Jonah
• Primary Examiner(s): Leland, III, Edwin S
• Application Number:
  US15/425,099
• Publication Number:
  US 20180182381A1
• Time in Patent Office: 834 Days
• Field of Search: 704 9
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: G06F 16/29 Geographical information databases
• CPC Class Codes: G06F 16/9537 Spatial or temporal dependent retrieval, e.g. spatiotemporal queries
• CPC Class Codes: G06F 40/289 Phrasal analysis, e.g. finite state techniques or chunking
• CPC Class Codes: G06F 40/30 Semantic analysis
• CPC Class Codes: G06Q 30/02 Marketing; Price estimation or determination; Fundraising
• CPC Class Codes: G06Q 30/0201 Market modelling; Market analysis; Collecting market data
• CPC Class Codes: G06Q 30/0202 Market predictions or forecasting for commercial activities
• CPC Class Codes: G06Q 30/0241 Advertisements
• CPC Class Codes: G06Q 30/0244 Optimization
• CPC Class Codes: G06Q 30/0273 Determination of fees for advertising
• CPC Class Codes: G06Q 30/0275 Auctions

Abstract

An accurate thought map is created by recording people'"'"'s many utterances of natural language expressions together with the location at which each expression was made. The expressions are input into a Natural Language Understanding system including a semantic parser, and the resulting interpretations stored in a database with the geolocation of the speaker. Emotions, concepts, time, user identification, and other interesting information may also be detected and stored. Interpretations of related expressions may be linked in the database. The database may be indexed and filtered according to multiple aspects of interpretations such as geolocation ranges, time ranges or other criteria, and analyzed according to multiple algorithms. The analyzed results may be used to render map displays, determine effective locations for advertisements, preemptively fetch information for users of mobile devices, and predict the behavior of individuals and groups of people.

Claims

1. A method of predicting a person'"'"'s behavior, the method comprising:

storing, in a database, interpretations of prior natural language expressions obtained from a user device and one or more geolocation based movement behaviors of the user device that occurred within a specific amount of time after the prior natural language expressions were received, each of the stored one or more geolocation based movement behaviors being related to one or more of the stored interpretations of the prior natural language expressions;

receiving a current natural language expression after the one or more geolocation based movement behaviors occurred;

interpreting the current natural language expression to create a current interpretation;

comparing the current interpretation to the stored interpretations of the prior natural language expressions to identify a closest matching stored interpretation of the stored interpretations;

predicting the person'"'"'s geolocation based movement behavior by selecting a stored geolocation based movement behavior, of the stored one or more geolocation based movement behaviors, related to the closest matching stored interpretation; and

providing the person'"'"'s predicted geolocation based movement to an entity.

2. The method of claim 1, further comprising:

storing, in the database, user IDs in association with the stored interpretations of the prior natural language expressions; and

filtering the database according to a current user ID of a user that uttered the current natural language expression,

wherein the filtering selects, from among the stored interpretations, interpretations that are associated with user IDs that match the current user ID.

3. The method of claim 1, wherein the prior natural language expressions are unrelated to an operational function of the user device.

4. At least one non-transitory computer readable medium storing code that, if executed by at least one computer processor would cause the at least one computer processor to:

store, in a database, interpretations of prior natural language expressions obtained from a user device and one or more geolocation based movement behaviors of the user device that occurred within a specific amount of time after the prior natural language expressions were received, each of the stored one or more geolocation based movement behaviors being related to one or more of the stored interpretations of the prior natural language expressions;

receive a current natural language expression after the one or more geolocation based movement behaviors occurred;

interpret the current natural language expression to create a current interpretation;

compare the current interpretation to the stored interpretations of the prior natural language expressions to identify a closest matching stored interpretation of the stored interpretations;

predict the person'"'"'s geolocation based movement behavior by selecting a stored geolocation based movement behavior, of the stored one or more geolocation based movement behaviors, related to the closest matching stored interpretation; and

provide the person'"'"'s predicted geolocation based movement to an entity.

5. The at least one non-transitory computer readable medium of claim 4, wherein the code, if executed by at least one computer processor further causes the at least one computer processor to:

store, in the database, user IDs in association with the stored interpretations of the prior natural language expressions; and

filter the database according to a current user ID of a user that uttered the current natural language expression, by selecting, from among the stored interpretations, interpretations that are associated with user IDs that match the current user ID.

6. The at least one non-transitory computer readable medium of claim 4, wherein the prior natural language expressions are unrelated to an operational function of the user device.

7. A system including one or more processors coupled to memory, the memory loaded with computer instructions to predict a person'"'"'s behavior, the instructions, when executed on the processors, implement actions comprising:

storing, in a database, interpretations of prior natural language expressions obtained from a user device and one or more geolocation based movement behaviors of the user device that occurred within a specific amount of time after the prior natural language expressions were received, each of the stored one or more geolocation based movement behaviors being related to one or more of the stored interpretations of the prior natural language expressions;

receiving a current natural language expression after the one or more geolocation based movement behaviors occurred;

interpreting the current natural language expression to create a current interpretation;

comparing the current interpretation to the stored interpretations of the prior natural language expressions to identify a closest matching stored interpretation of the stored interpretations;

predicting the person'"'"'s geolocation based movement behavior by selecting a stored geolocation based movement behavior, of the stored one or more geolocation based movement behaviors, related to the closest matching stored interpretation; and

providing the person'"'"'s predicted geolocation based movement to an entity.