Collaborator discovery method and system

• US 6,681,247 B1
• Filed: 10/18/1999
• Issued: 01/20/2004
• Est. Priority Date: 10/18/1999
• Status: Expired due to Term

• Current Assignee: HRL Laboratories LLC (The Boeing Co.)
• Original Assignee: HRL Laboratories LLC (The Boeing Co.)
• Inventors: Payton, David W.
• Primary Examiner(s): Dinh, Dung C.
• Assistant Examiner(s): PARTON, KEVIN S
• Application Number:
  US09/420,041
• Time in Patent Office: 1,555 Days
• Field of Search: 707/6, 707/10, 707/1, 707/5, 709/219, 709/218, 709/217, 709/204, 709/206, 709/203
• US Class Current: 709/217
• CPC Class Codes: G06F 16/284 Relational databases
• CPC Class Codes: Y10S 707/99935 Query augmenting and refining, e.g. inexact access
• CPC Class Codes: Y10S 707/99936 Pattern matching access

Abstract

A collaborator discovery method and system is presented for tracking and correlating user activities with respect to information resources on an electronic network to assist users in finding others with common interests. The system includes a monitor for tracking user activities, an entry processor for updating and providing monitored activities to a match database, and a matcher to correlate user activities and to diffuse user interests to information resources that have not yet been visited. Long-term and short-term user interests are tracked, and are decayed over time for resources not recently visited. Information sources whose associated interest level has sufficiently decayed are pruned, and their entries in the match database are eliminated. An interactive messaging system is also provided for users to interact, while preserving their anonymity.

Claims

1. A method for collaborator discovery among a plurality of users including the steps of:

(a) providing a user history including a plurality of entries, with each entry including a user identity associated with each particular user and a reference to a particular item accessed by that user;

(b) associating particular items in the user history by providing a measure of similarity between the particular items;

(c) uniquely associating at least one scent score to each particular item accessed by a particular user;

(d) diffusing the at least one scent score associated with a particular item accessed by a particular user to another item by generating at least one diffusion scent score from the combination of the measure of similarity between the particular item and the other item and the at least one scent score, and incorporating the at least one diffusion scent score into the at least one scent score of the other item;

(e) repeating step (d) for all items which have at least one scent score; and

(f) determining scent match scores by correlating the scent scores from all of the particular items to find users with common interests.

2. A method for collaborator discovery among a plurality of users as set forth in claim 1, wherein the user history is generated by monitoring and recording the real-time accesses of the plurality of users, and wherein steps (b) through (f) of claim 1 are repeated a plurality of times to provide a continual update of the scent scores.

3. A method for collaborator discovery among a plurality of users as set forth in claim 1, further including the step of filtering the items in the user history using a criteria to eliminate undesirable items from the user history.

4. A method for collaborator discovery among a plurality of users as set forth in claim 1, wherein the measure of similarity between the particular items associated in step (b) is based on the temporal proximity of access between the particular items.

5. A method for collaborator discovery among a plurality of users as set forth in claim 1, wherein the scent scores are represented by scalar values, and wherein a step of increasing the scent scores in proportion to the number of times a particular item is accessed by a particular user is added between steps (c) and (d).

6. A method for collaborator discovery among a plurality of users as set forth in claim 1, further including the step of providing each of the plurality of users with information regarding the correlation of their scent scores with the scent scores of others of the plurality of users after step (f).

7. A method for collaborator discovery among a plurality of users as set forth in claim 1, wherein the plurality of users are provided with anonymous identifications.

8. A method for collaborator discovery among a plurality of users as set forth in claim 1, wherein each of the plurality of users is provided a method for messaging to allow interaction between the plurality of users.

9. A method for collaborator discovery among a plurality of users as set forth in claim 2, wherein items accessed include Internet web pages.

10. A method for collaborator discovery among a plurality of users as set forth in claim 2, wherein the scent scores are decayed over time.

11. A method for collaborator discovery among a plurality of users as set forth in claim 2, wherein the at least one scent score for each particular user and information item includes a short-term scent score and a long-term scent score, and where, for each subsequent access of each particular item by a particular user, the short-term scent score and long-term scent score are increased in proportion to the number of accesses by the particular user such that the short-term scent score increases more rapidly than the long-term scent score.

12. A method for collaborator discovery among a plurality of users as set forth in claim 5, wherein a maximum scent score value is set such that when a particular scent score reaches the maximum scent score value, it ceases to increase.

13. A method for collaborator discovery among a plurality of users as set forth in claim 12, wherein the long-term scent scores and the short-term scent scores are decayed over time with a decay rate such that the long-term scent scores are decayed more slowly than the short-term scent scores.

14. A method for collaborator discovery among a plurality of users as set forth in claim 13,

a. wherein the short-term scent score and long-term scent scores are associated with each particular user according to the following,

15. A method for collaborator discovery among a plurality of users as set forth in claim 14, wherein the item from which the scent score is diffused is identified as a source item A and the item to which the scent score is diffused is identified as a destination item B, and the scent score diffusion is performed according to,

16. A method for collaborator discovery among a plurality of users as set forth in claim 15, wherein the correlation of the scent scores between user a, representing a particular one of the plurality of users, and user b, representing another of the plurality of users, where item p represents a particular one of the plurality of items, is performed by the following, ${\mathrm{SS}}_{}$

_Matchab=∑

p

SSap×

SSbpStotp∑

p

SSap2

∑

p

SSbp2,

SL

_Matchab=∑

p

SSap×

SLbpStotp∑

p

SSap2

∑

p

SLbp2, 



and${\mathrm{LL}<br><br>\mathrm{\_Match}}_{\mathrm{ab}}=\frac{\underset{p}{\sum <br><br>}<br><br>\frac{{\mathrm{SL}}_{\mathrm{ap}}\times <br><br>{\mathrm{SL}}_{\mathrm{bp}}}{{\mathrm{Stot}}_p}}{\sqrt{\underset{p}{\sum <br><br>}<br><br>{\mathrm{SL}}_{\mathrm{ap}}^2}<br><br>\sqrt{\underset{p}{\sum <br><br>}<br><br>{\mathrm{SL}}_{\mathrm{bp}}^2}},\text{\hspace{1em}<br><br>}<br><br>\mathrm{where}$

SS_Match_ab is the match between short-term scent scores of user a and user b;

SL_Match_ab is the match between the short-term scent score of user a and the long-term scent score of user b;

LL_Match_ab is the match between the long-term scent scores of users a and b;

Stot_p is the total number of distinct user scent scores that can be found at item p;

SS_ap is the short-term scent score assigned to user a at item p; and

SL_ap is the long-term scent score assigned to user a at item p.

17. A system for collaborator discovery among a plurality of users including:

a. an activity monitor which provides a user history, said user history including a plurality of entries, with the plurality of entries including a user identity associated with each particular user and a reference to a particular item accessed by that user;

b. an entry processor connected to the activity monitor to receive the plurality of entries of the user history from the activity monitor, said entry processor operative to associate pairs of particular items in the user history to provide a measure of similarity for each pair of particular items, and to uniquely associate at least one scent score for each particular item accessed by a particular user;

c. a match database connected to the entry processor to receive and store the measure of similarity and the at least one scent score;

d. a matcher connected to the match database to receive the measure of similarity and the at least one scent score, and to diffuse the at least one scent score to other particular items in the user history in proportion to the measure of similarity and to correlate the scent scores of all of the particular items in the user history to determine users with common interests.

18. A system for collaborator discovery among a plurality of users as set forth in claim 17, wherein the user history provided by the activity monitor is generated from information regarding the real-time activities of users with respect to a plurality of items.

19. A system for collaborator discovery among a plurality of users as set forth in claim 17, further including means for providing user anonymity.

20. A system for collaborator discovery among a plurality of users as set forth in claim 17, further including means to allow for interaction between users.

21. A system for collaborator discovery among a plurality of users as set forth in claim 17, wherein the entry processor includes an information item type filter operative to eliminate unimportant entries upon receipt from the activity monitor to provide a plurality of filtered entries.

22. A system for collaborator discovery among a plurality of users as set forth in claim 17, wherein,

a. the entry processor further includes an item association engine, said item association engine linked to the information item type filter to receive the plurality of filtered entries therefrom and to assign a measure of similarity for each pair of particular items; and

b. the match database further includes means to receive and store the measure of similarity for each pair of particular items from the entry processor.

23. A system for collaborator discovery among a plurality of users as set forth in claim 17, wherein the users of the system access the system by computer and wherein the activity monitor is distributed across the computers used by the users.

24. A system for collaborator discovery among a plurality of users as set forth in claim 17, wherein the activity monitor is centralized.

25. A system for collaborator discovery among a plurality of users as set forth in claim 18, wherein,

a. the entry processor further includes a scent update engine which receives the plurality of filtered entries from the filter and uniquely associates at least one scent score for each particular item accessed by a particular user;

and

b. the match database further includes means to receive and store the at least one scent score for each particular item accessed by a particular user from the scent update engine.

26. A system for collaborator discovery among a plurality of users as set forth in claim 18, wherein the matcher further includes a diffusion engine linked to the match database to receive the measure of similarity for each pair of particular items and to receive the scent score corresponding to at least one particular item of the pair of particular items for which the measures of similarity were received, and further to utilize the measures of similarity and the at least one scent score to diffuse the at least one scent score from one item of the pair of particular items to the other item of the pair of particular items to generate a diffused scent score, and to incorporate the diffused scent score into the match database.

27. A system for collaborator discovery among a plurality of users as set forth in claim 17, wherein the scent scores are increased in proportion to the number of times a particular item is accessed by a particular user.

28. A system for collaborator discovery among a plurality of users as set forth in claim 19, wherein the users of the system access the system by computer and wherein the means for providing user anonymity is distributed across the computers used by the users.

29. A system for collaborator discovery among a plurality of users as set forth in claim 19, wherein the means for providing user anonymity is centralized.

30. A system for collaborator discovery among a plurality of users as set forth in claim 20, wherein the means to allow for interaction between users is a chat system.

31. A system for collaborator discovery among a plurality of users as set forth in claim 20, wherein the means to allow for interaction between users is an e-mail system.

32. A system for collaborator discovery among a plurality of users as set forth in claim 27, wherein the matcher further includes a decay engine linked to the match database decay the measure of similarity and the at least one scent score for each particular item for each user over time.

33. A system for collaborator discovery among a plurality of users as set forth in claim 27, wherein the at least one scent score associated for each particular user and information item by the entry processor includes a short-term scent score and a long-term scent score where, for each subsequent access of each particular item by a particular user, the entry processor increases the short-term scent score and long-term scent score in proportion to the number of accesses of a particular item by the particular user such that the short-term scent score increases more rapidly than the long-term scent score.

34. A system for collaborator discovery among a plurality of users as set forth in claim 27, wherein a maximum scent score value is set such that when a particular scent score reaches the maximum scent score value, it ceases to increase.

35. A system for collaborator discovery among a plurality of users as set forth in claim 28, wherein the decay engine includes a decay rate used to decay the long-term scent scores and the short-term scent scores over time with the decay rate chosen such that the long-term scent scores decay more slowly than the short-term scent scores.

36. A system for collaborator discovery among a plurality of users as set forth in claim 35,

a. wherein the entry processor associates short-term scent score and long-term scent scores with each particular user according to the following,

 

SS=CS

37. A system for collaborator discovery among a plurality of users as set forth in claim 36, wherein the item from which the scent score is diffused is identified as a source item A and the item to which the scent score is diffused is identified as a destination item B, and the scent score diffusion is performed by the diffusion engine according to,

38. A system for collaborator discovery among a plurality of users as set forth in claim 37, wherein the matcher correlates the scent scores between user a, representing a particular one of the plurality of users, and user b, representing another of the plurality of users, where item p represents a particular one of the plurality of items, according to the following, ${\mathrm{SS}}_{}$

_Matchab=∑

p

SSap×

SSbpStotp∑

p

SSap2

∑

p

SSbp2,

SL

_Matchab=∑

p

SSap×

SLbpStotp∑

p

SSap2

∑

p

SLbp2, 



and${\mathrm{LL}<br><br>\mathrm{\_Match}}_{\mathrm{ab}}=\frac{\underset{p}{\sum <br><br>}<br><br>\frac{{\mathrm{SL}}_{\mathrm{ap}}\times <br><br>{\mathrm{SL}}_{\mathrm{bp}}}{{\mathrm{Stot}}_p}}{\sqrt{\underset{p}{\sum <br><br>}<br><br>{\mathrm{SL}}_{\mathrm{ap}}^2}<br><br>\sqrt{\underset{p}{\sum <br><br>}<br><br>{\mathrm{SL}}_{\mathrm{bp}}^2}},\text{\hspace{1em}<br><br>}<br><br>\mathrm{where}$

SS_Match_ab is the match between short-term scent scores of user a and user b;

SL_Match_ab is the match between the short-term scent score of user a and the long-term scent score of user b;

LL_Match_ab is the match between the long-term scent scores of users a and b;

Stot_p is the total number of distinct user scent scores that can be found at item p;

SS_ap is the short-term scent score assigned to user a at item p; and

SL_ap is the long-term scent score assigned to user a at item p.