Determining and scoring task indications

• US 9,436,755 B1
• Filed: 01/26/2014
• Issued: 09/06/2016
• Est. Priority Date: 01/26/2014
• Status: Active Grant

• Current Assignee: Google LLC (Alphabet Inc.)
• Original Assignee: Google Inc. (Alphabet Inc.)
• Inventors: Orr, Dave, Subramanya, Amarnag, Pellicer, Edgar Gonzalez, Agarwal, Shalini
• Primary Examiner(s): Filipczyk, Marc
• Application Number:
  US14/164,232
• Time in Patent Office: 954 Days
• Field of Search: 704/257, 707737-738, 707/750
• US Class Current: 1/1
• CPC Class Codes: G06F 16/3322 using system suggestions G06F16/3325 takes precedence
• CPC Class Codes: G06F 16/3344 using natural language analysis
• CPC Class Codes: G06F 40/30 Semantic analysis

Abstract

Methods and apparatus related to determining and scoring task indications. For example, methods and apparatus may determine a plurality of interrogative sentences in a plurality of messages, determine starting n-grams of the interrogative sentences, determine task indications based on the starting n-grams, and determine task association scores for the task indications. Each task indication may be determined based on a set of one or more starting n-grams that share similarities. One or more individuals and/or applications may assign one or more task association measures to a task indication based on likelihood that the task indication is associated with a task request. The task association score for the task indication may be determined based on the task association measures that are assigned to the task indication.

Claims

1. A method performed by one or more processors based on instructions stored in memory, comprising:

identifying, by one or more of the processors based on the instructions stored in the memory, a plurality of messages from one or more databases;

determining, by one or more of the processors based on the instructions stored in the memory, a plurality of interrogative sentences from the messages;

determining, by one or more of the processors based on the instructions stored in the memory, starting n-grams from the interrogative sentences;

determining, by one or more of the processors based on the instructions stored in the memory, a group of task indications, wherein each of the task indications of the group is based on a set of one or more of the starting n-grams, and wherein determining to include a given task indication of the task indications in the group is based on a count of the starting n-grams that conform to the given task indication;

providing, by one or more of the processors based on the instructions stored in the memory, the group of the task indications;

receiving, by one or more of the processors based on the instructions stored in the memory and in response to providing the group of the task indications, one or more task association measures for each of the task indications of the group, wherein a given task association measure for a given task indication is indicative of likelihood that the given task indication is associated with a task request;

determining, by one or more of the processors based on the instructions stored in the memory, a task association score for each of the task indications of the group, wherein the task association score for the given task indication is based on the one or more task association measures received for the given task indication; and

storing, by one or more of the processors based on the instructions stored in the memory, the task association score for each of a plurality of the task indications of the group.

2. The method of claim 1, where providing the group of the task indications includes providing the group of the task indications, independent of other information of the interrogative sentences, to one or more computing devices for assignment of the task association measures by one or more individuals.

3. The method of claim 1, wherein determining the starting n-gram for a given interrogative sentence of the interrogative sentences includes identifying a verb in the given interrogative sentence and determining the starting n-gram based on one or more n-grams that occur before the verb in the given interrogative sentence.

4. The method of claim 3, wherein the starting n-gram for the given interrogative sentence includes the verb and one or more of the n-grams that occur before the verb in the given interrogative sentence.

5. The method of claim 1, wherein determining the plurality of interrogative sentences includes determining the interrogative sentences based on at least one n-gram of the interrogative sentences, wherein the at least one n-gram is indicative of an interrogative sentence.

6. The method of claim 5, wherein the at least one n-gram includes a punctuation mark.

7. The method of claim 6, wherein the punctuation mark is a question mark.

8. The method of claim 1, wherein each of the task association measures is a binary value.

9. The method of claim 1, further comprising:

associating, in a database, one or more of the task indications with a corresponding said task association score.

10. The method of claim 1, further including:

determining the task association score for the given task indication satisfies a threshold task association score indicative of a likely task request; and

storing, in a database, the given task indication as a likely task indication based on determining the task association score for the given task indication satisfies the threshold task association score.

11. The method of claim 1, further comprising:

determining a subset of the task indications with corresponding said task association scores that satisfy a threshold task association score indicative of a likely task request; and

storing the subset of the task indications as likely task indications.

12. The method of claim 11, wherein storing the subset of the task indications as likely task indications includes storing, in a database, each of the task indications of the subset and associated said task association scores.

13. The method of claim 1, wherein identifying the plurality of messages includes identifying the messages based on senders of the messages.

14. The method of claim 13, wherein identifying the plurality of messages based on the senders includes restricting the messages to those indicated as having a sender who is an individual.

15. The method of claim 1, wherein identifying the plurality of messages includes identifying the messages based on recipients of the messages.

16. The method of claim 1, wherein each of the starting n-grams is only a portion of a respective one of the interrogative sentences.

17. A system including memory and one or more processors operable to execute instructions stored in memory, comprising instructions to:

identify a plurality of messages from one or more databases;

determine a plurality of interrogative sentences from the messages;

determine starting n-grams from the interrogative sentences;

determine a group of task indications, wherein each of the task indications of the group is based on a set of one or more of the starting n-grams, and wherein determining to include a given task indication of the task indications in the group is based on a count of the starting n-grams that conform to the given task indication;

provide the group of the task indications;

receive, in response to providing the group of the task indications, one or more task association measures for each of the task indications of the group, wherein a given task association measure for a given task indication is indicative of likelihood that the given task indication is associated with a task request;

determine a task association score for each of the task indications of the group, wherein the task association score for the given task indication is based on the one or more task association measures received for the given task indication; and

store the task association score for each of a plurality of the task indications of the group.

18. The system of claim 17, where the instructions to provide the group of the task indications includes instructions to provide the group of the task indications, independent of other information of the interrogative sentences, to one or more computing devices for assignment of the task association measures by one or more individuals.

19. The system of claim 17, wherein the instructions to determine the plurality of interrogative sentences includes instructions to determine the interrogative sentences based on at least one n-gram of the interrogative sentences, wherein the at least one n-gram is indicative of an interrogative sentence.

20. The system of claim 17, wherein the instructions to identify the plurality of messages includes instructions to identify the messages based on senders of the messages.

21. The system of claim 20, wherein the instructions to identify the plurality of messages based on the senders includes instructions to restrict the messages to those indicated as having a sender who is an individual.

22. The system of claim 17, wherein the instructions to identify the plurality of messages includes instructions to identify the messages based on recipients of the messages.

23. A non-transitory computer readable storage medium storing at least one program configured for execution by at least one processor of a computer system, the at least one program comprising instructions to:

identify a plurality of messages from one or more databases;

determine a plurality of interrogative sentences from the messages;

determine starting n-grams from the interrogative sentences;

determine a group of task indications, wherein each of the task indications of the group is based on a set of one or more of the starting n-grams, and wherein determining to include a given task indication of the task indications in the group is based on a count of the starting n-grams that conform to the given task indication;

provide the group of the task indications;

receive, in response to providing the group of the task indications, one or more task association measures for each of the task indications of the group, wherein a given task association measure for a given task indication is indicative of likelihood that the given task indication is associated with a task request;

determine a task association score for each of the task indications of the group, wherein the task association score for the given task indication is based on the one or more task association measures received for the given task indication; and

store the task association score for each of a plurality of the task indications of the group.