Methods and systems for dynamically suggesting answers to questions submitted to a portal of an online service

• US 8,595,207 B2
• Filed: 05/19/2011
• Issued: 11/26/2013
• Est. Priority Date: 06/14/2010
• Status: Active Grant

• Current Assignee: Salesforce.com, Inc.
• Original Assignee: Salesforce.com, Inc.
• Inventors: Mathrubootham, Ganesh, McFarlane, Patrick R., Yang, Suchun
• Primary Examiner(s): Boccio, Vincent F
• Application Number:
  US13/111,195
• Publication Number:
  US 20110307498A1
• Time in Patent Office: 922 Days
• Field of Search: 707/705, 707/721, 707/736, 707/748, 707/765, 707/766, 707/767, 707/768, 707/758, 707/759
• US Class Current: 707/707
• CPC Class Codes: G06F 16/24578 using ranking
• CPC Class Codes: G06F 16/3322 using system suggestions G06F16/3325 takes precedence
• CPC Class Codes: G06F 16/334 Query execution G06F16/335 takes precedence
• 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 16/9538 Presentation of query results

Abstract

In accordance with embodiments, there are provided mechanisms and methods for dynamically suggesting answers to questions submitted to a portal of an online service. These mechanisms and methods for dynamically suggesting answers to questions submitted to a portal of an online service can enable embodiments to provide help document suggestions with limited user input. The ability of embodiments to provide help document search term suggestions with limited user input can enable users to quickly and efficiently find useful help documents.

Claims

1. A method for dynamically providing search suggestions to a user of an on-demand service in

a multi-tenant database environment, the method comprising:

receiving, at a host system that is remote from a user system, user input for conducting a search, the host system having a processor system including at least one processor and a memory system;

regardless of whether the user input received so far is all of the input the user intended to enter, if the user input entered so far includes more characters than a predetermined number of characters, automatically searching, by the processor system, a storage area in the memory system at the host system for stored search terms from prior searches that are similar to the user input;

automatically determining, by the host system, a score for the search terms from a set of factors;

the automatically determining including

capping a value for a given score by determining whether the value is greater than a predetermined threshold, and if the value is greater than the predetermined threshold, setting the score to the predetermined threshold,

normalizing the given score by dividing one or more values that determine the score by a normalization value, therein as a result of the normalization and the capping, the score never exceeds a predetermined value,

automatically determining, by the host, a ranking for the search terms found based on the score,

the ranking being a value that indicates an expected likelihood that the user will be interested in the search term,

sending, from the host system to the user system, the search terms found listed in order of ranking as search suggestions completing the user input.

2. The method of claim 1, further comprising

receiving, at the host system, a search term for conducting a search;

in response, at the host system,

conducting a search for documents having the search term, and

automatically incrementing a value representing a number of votes for the search term of the search;

saving the incremented value in association with the search term.

3. The method of claim 1, wherein the search suggestions that are sent are formatted such that search suggestions having a higher ranking are displayed higher than search suggestions having a lower ranking.

4. The method of claim 1, wherein the search suggestions that are sent, are previously submitted search terms.

5. The method of claim 1, further comprising:

storing in a first object for storing data, search suggestions in association with scores;

storing in a second object for storing data, raw search terms that were previously used for conducting a search, the raw search terms being stored in association with a value representing an indication as to how many times users entered each of the raw search terms.

6. The method of claim 1, wherein the user input comprises at least four characters. and the searching, the determining, and the sending are performed after receiving the at least four characters of the user input regardless of whether the user is finished entering all characters of the user input.

7. The method of claim 1, further comprising:

receiving, at the host system, a search term for conducting a search;

updating, a ranking of search terms, based on the calculated score;

conducting, a search based on the search term; and

sending, search results based on the search conducted, to the user system.

8. The method of claim 7, wherein the updating of the ranking of search terms based on the

calculated score results in search terms with higher calculated scores ranking higher, and placing search terms with a higher ranking higher in the order of the ranking listed in the search suggestions sent to the user machine.

9. The method of claim 7, wherein the predetermined set of factors for calculating the score includes

a number of previous search occurrences based on the search term,

a top relevancy of search results based on the search term,

an average relevancy of search results based on the search term,

a range of relevancy of search results based on the search term, and

a number of words of the search term found in the search term.

10. The method of claim 7, wherein the normalizing includes normalizing the set of factors for calculating a score, the calculating of a score for a search suggestion includes

weighting the set of normalized factors for calculating a score, and

summing the weighted set of normalized factors to determine the score.

11. The method of claim 7, the normalizing includes normalizing the set of factors for calculating a score, the method further comprising:

storing in a first object for storing data, search suggestions in association with a score, wherein the search suggestions are previously submitted search terms;

storing in a second object for storing data, raw search terms for conducting a search that were previously used in a search; and

the calculating of the score for a search suggestion includes

weighting the set of normalized factors for calculating a score, and

summing the weighted set of normalized factors to determine the score.

12. The method of claim 7, the method further comprising:

storing in a first object for storing data, search suggestions in association with a score, wherein the search suggestions are previously submitted search terms;

storing in a second object for storing data, raw search terms that were previously used for conducting a search; and

the set of factors for calculating of the score for a search suggestion includes

a number of previous search occurrences based on the search term, top relevancy of search results based on the search term,

an average relevancy of search results based on the search term,

a range of relevancy of search results based on the search term, and

a number of words of the search term found in the search term;

the calculating of the score for a search suggestion includes

weighting the set of normalized factors for calculating a score, and

summing the weighted set of normalized factors to determine the score.

13. The method of 5, the method further comprising:

determining whether the value of the indication is greater than a threshold;

if the value of the indication is greater than the threshold copying the raw search term in second object and placing the raw search term in the first object as a search suggestion.

14. The method of 5, further comprising:

grouping together raw search terms that are in the second object into a group, based on a degree of similarity;

determining a score for the group based on how often a search term from the group was submitted by users;

determining whether to copy the group into the first object based on the score of the group.

15. A non-transitory computer-readable medium storing one or more sequences of instructions for causing one

or more processors to implement a method for dynamically providing search suggestions to a

user of an on-demand service in a multi-tenant database environment, the method comprising the steps of:

receiving, at a host system that is remote from a user system, user input for conducting a search;

sending, from the host system to the user system, search suggestions based on the user input;

receiving, at the host system, a search term for conducting a search;

calculating, a score for the search term for conducting a search, based on a set of factors;

updating, a ranking of search terms, based on the calculated score;

conducting, a search based on the search term; and

sending, search results based on the search conducted, to the user system; and

the calculating of the score for the search suggestion includes

normalizing a set of factors for calculating a score, the set of factors for calculating a score includes

a number of previous search occurrences based on the search term,

a top relevancy of search results based on the search term,

an average relevancy of search results based on the search term, and

a number of words of the search term found in the search suggestion;

weighting the set of normalized factors for calculating a score, and

summing the weighted set of normalized factors to determine the score.

16. The non-transitory computer-readable medium of claim 15, the method further comprising:

storing in a first object for storing data, ranked search suggestions, wherein the search suggestions are previously submitted search terms;

storing in a second object for storing data, search terms for conducting a search; and

calculating a score for a search suggestion includes

normalizing the set of factors for calculating a score,

weighting the set of normalized factors for calculating a score, and

summing the weighted set of normalized factors to determine the score.

17. A machine resident host system for dynamically providing search suggestions to a user of an

on-demand service in a multi-tenant database environment, the host system comprising:

a processor system,

a memory system including at least volatile memory, and non-volatile memory;

the non-volatile memory including at least one machine readable medium carrying one or more sequences of instructions which when implemented causes the processor system to implement a method comprising;

receiving, at a host system that is remote from a user system, user input for conducting a search;

sending, from the host system to the user system, search suggestions based on the user input;

receiving, at the host system, a search term for conducting a search;

calculating, by the a processor system of the host system, a score for search suggestions related to the search term for conducting a search, based on a set of factors, for each given search suggestion, one of the factors of the score of the given search suggestion being a range of relevancy of the documents retrieved as a result of a search conducted with that given search suggestion as a search query;

updating, by the host system, a ranking of search terms, based on the calculated score;

conducting, a search based on the search term; and

sending, search results based on the search conducted, to the user system.

18. The machine resident host system of claim 17, the method further comprising:

storing in a first object for storing data, ranked search suggestions, wherein the search suggestions are previously submitted search terms;

storing in a second object for storing data, search terms for conducting a search; and

the calculating of the score for a search suggestion includes

normalizing the set of factors for calculating a score,

weighting the set of normalized factors for calculating a score, and

summing the weighted set of normalized factors to determine the score.

19. The machine resident host system of claim 17, the method further comprising:

storing in a first object for storing data, ranked search suggestions, wherein the search suggestions are previously submitted search terms;

storing in a second object for storing data, search terms for conducting a search; and

calculating a score for a search suggestion includes

normalizing the set of factors for calculating a score, the set of factors for calculating a score includes

a number of previous search occurrences based on the search term,

a top relevancy of search results based on the search term,

an average relevancy of search results based on the search term,

a range of relevancy of search results based on the search term, and

a number of words of the search term found in the search term,

weighting the set of normalized factors for calculating a score, and

summing the weighted set of normalized factors to determine the score.