COMPUTER IMPLEMENTED METHODS AND APPARATUS FOR IDENTIFYING A TOPIC FOR A TEXT

• US 20140067814A1
• Filed: 09/04/2013
• Published: 03/06/2014
• Est. Priority Date: 09/04/2012
• Status: Active Grant

• Current Assignee: Salesforce.com, Inc.
• Original Assignee: Salesforce.com, Inc.
• Inventors: Palmert, Joel, Gustafson, Erik
• Granted Patent: US 9,292,589 B2
• Time in Patent Office: Days
• US Class Current: 707/738
• CPC Class Codes: G06F 16/285 Clustering or classification
• CPC Class Codes: G06F 16/35 Clustering; Classification
• CPC Class Codes: G06F 16/951 Indexing; Web crawling techniques
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: G06F 16/9536 Search customisation based on social or collaborative filtering

Abstract

Disclosed are methods, apparatus, systems, and computer-readable storage media for identifying a topic for a text. In some implementations, one or more servers maintain a plurality of data entries in one or more database tables storing text data, each data entry of a first portion of the data entries including: a text sequence, a topic, and a text-to-topic association score indicating a number of times that the text sequence appears in a processed text associated with the topic, each data entry of a second portion of the data entries including a total word score indicating a number of times that a respective text sequence appears in one or more processed texts. The one or more servers may receive an incoming text and identify a topic for the incoming text by processing the text sequences of the incoming text in relation to the data entries in the database tables.

Claims

1. A computer implemented method for identifying a topic for a text, the method comprising:

maintaining, by one or more servers, a plurality of data entries in one or more database tables storing text data, each data entry of a first portion of the data entries including;

a text sequence, a topic, and a text-to-topic association score indicating a number of times that the text sequence appears in a processed text associated with the topic, each data entry of a second portion of the data entries including a total word score indicating a number of times that a respective text sequence appears in one or more processed texts;

receiving an incoming text having a length and including one or more text sequences; and

identifying a topic for the incoming text by processing the one or more text sequences of the incoming text in relation to the data entries in the one or more database tables.

2. The method of claim 1, wherein the text sequence includes one or more words.

3. The method of claim 1, wherein identifying the topic for the incoming text comprises:

for each text sequence of the incoming text, generating a subtotal topic score for the text sequence and the topic;

determining a total topic score for the incoming text and the topic by summing the subtotal topic scores, the total topic score indicating relevance of the topic to the incoming text; and

determining that the total topic score for the incoming text meets or traverses a threshold.

4. The method of claim 3, further comprising:

comparing the total topic score with other total topic scores for other topics; and

ranking the total topic score with the other total topic scores in order of relevance.

5. The method of claim 3, wherein generating the subtotal topic score for the text sequence and the topic comprises:

identifying a first data entry containing the text sequence and the topic;

normalizing the text-to-topic association score of the identified first data entry and a total word score associated with the text sequence;

dividing the normalized text-to-topic association score by a first function of the normalized total word score to generate an intermediate score; and

dividing the intermediate score by a second function of the length of the incoming text to generate a subtotal topic score.

6. The method of claim 5, wherein normalizing the text-to-topic association score and the total word score comprises:

dividing the text-to-topic association score by an inflation factor; and

dividing the total word score by the inflation factor.

7. The method of claim 6, wherein the inflation factor has a value based on a measure of time.

8. The method of claim 7, wherein the inflation factor is an exponential function of a measure of time.

9. The method of claim 5, wherein the first function and the second function are a square root.

10. The method of claim 1, the method further comprising:

transmitting the identified topic to a computing device for display as a suggested topic for the incoming text.

11. The method of claim 10, wherein the identified topic is transmitted to the computing device with one or more other suggested topics.

12. The method of claim 1, the method further comprising:

responsive to a request to assign a topic to the incoming text, updating the one or more database tables with the one or more text sequences of the incoming text and the requested topic.

13. The method of claim 12, wherein updating the one or more database tables comprises:

for each text sequence of the incoming text;

identifying or creating a first data entry of a first database table that includes the text sequence and the identified topic;

incrementing the text-to-topic association score of the identified first data entry by an inflation factor;

identifying or creating a second data entry of a second database table that includes the text sequence; and

incrementing the total word score of the identified second data entry by the inflation factor.

14. The method of claim 1, further comprising:

providing a first database table configured to store the first data entries;

providing a second database table configured to store text sequences and corresponding total word scores;

identifying one or more texts associated with one or more topics, each text having a plurality of text sequences;

for each identified text, updating the first and second database tables using the text sequences of the identified text and the topic associated with the identified text.

15. The method of claim 1, wherein a server identifies the topic for the incoming text in response to receiving from a computing device a request for suggested topics based on the incoming text.

16. The method of claim 15, wherein the request for suggested topics is generated by the computing device in response to a request to post the incoming text.

17. The method of claim 15, wherein the request for suggested topics is generated by the computing device in response to a request to select a topic to be assigned to the incoming text.

18. One or more computing devices for identifying a topic for a text, the one or more computing devices comprising:

one or more processors operable to execute one or more instructions to;

maintain a plurality of data entries in one or more database tables storing text data, each data entry of a first portion of the data entries including;

a text sequence, a topic, and a text-to-topic association score indicating a number of times that the text sequence appears in a processed text associated with the topic, each data entry of a second portion of the data entries including a total word score indicating a number of times that a respective text sequence appears in one or more processed texts;

receive an incoming text having a length and including one or more text sequences; and

identify a topic for the incoming text by processing the one or more text sequences of the incoming text in relation to the data entries in the one or more database tables.

19. A non-transitory computer-readable storage medium storing instructions executable by a server to perform a method for identifying a topic for a text, the method comprising:

maintaining, by one or more servers, a plurality of data entries in one or more database tables storing text data, each data entry of a first portion of the data entries including;

a text sequence, a topic, and a text-to-topic association score indicating a number of times that the text sequence appears in a processed text associated with the topic, each data entry of a second portion of the data entries including a total word score indicating a number of times that a respective text sequence appears in one or more processed texts;

receiving an incoming text having a length and including one or more text sequences; and

identifying a topic for the incoming text by processing the one or more text sequences of the incoming text in relation to the data entries in the one or more database tables.

20. The non-transitory computer-readable storage medium of claim 19, wherein identifying the topic for the incoming text comprises:

for each text sequence of the incoming text, generating a subtotal topic score for the text sequence and the topic;

determining a total topic score for the incoming text and the topic by summing the subtotal topic scores, the total topic score indicating relevance of the topic to the incoming text; and

determining that the total topic score for the incoming text meets or traverses a threshold.