Information routing

• US 6,266,667 B1
• Filed: 01/14/1999
• Issued: 07/24/2001
• Est. Priority Date: 01/15/1998
• Status: Expired due to Term

• Current Assignee: Telefonaktiebolaget LM Ericsson
• Original Assignee: Telefonaktiebolaget LM Ericsson
• Inventors: Olsson, Tomas
• Primary Examiner(s): Alam, Hosain T.
• Assistant Examiner(s): ALAM, SHAHID AL
• Application Number:
  US09/229,767
• Time in Patent Office: 922 Days
• Field of Search: 707/7, 707/10, 707/103, 707/104, 707/3, 707/5, 707/200, 707/201, 707/203, 05/52, 05/54, 05/77, 712/14, 713/176, 713/200, 379/112, 379/113, 379/201, 379/207, 379/220, 379/265, 379/93.15, 709/201, 709/203, 709/206, 709/228, 709/230
• US Class Current: 1/1
• CPC Class Codes: G06F 16/9535 Search customisation based on user profiles and personalisation
• CPC Class Codes: Y10S 707/99944 Object-oriented database structure
• CPC Class Codes: Y10S 707/99953 Recoverability

Abstract

The present invention relates to a problem with how to automatically or semi-automatically find and retrieve information for a user according to his preferences, is solved by sending an information object from a first interest agent to a second interest agent. Said information object comprising electronic information, a first array of references to other interest agents which recommends the electronic information and a second array of references to other interest agents which are indifferent to the content of the electronic information. Said second interest agent comprises a third array, where each position in said array comprises references to other interest agents and a confidence value. If the first agent which sent the information object is found in said third array and has a high confidence value the electronic information is presented before a user.

Claims

1. A method for finding and retrieving electronic information in a network, wherein:

a first interest agent sends an information object to a second interest agent;

said information object comprises electronic information, a first array of references to interest agents recommending said electronic information and a second array of references to interest agents which are indifferent to said electronic information;

said second interest agent comprises a third array where each position in said third array comprises at least a reference to an interest agent and a corresponding confidence value, and said second interest agent presents said electronic information to a user in dependence on the confidence value of the first interest agent;

the second interest agent further comprises a first model of interest, and when the confidence in said first interest agent is not high enough to present the information, the electronic information is presented if the electronic information in said information object and said first model of interest of the second interest agent match; and

the first model of interest comprises a first hash table, in which table a unique word has a unique position and where each position has a value indicating the importance of said position.

2. A method according to claim 1, wherein when the electronic information and said first model do not match, the electronic information is presented in dependence on a probability factor.

3. A method according to claim 1, wherein said second interest agent sends said information object to other interest agents if the user recommends the electronic information.

4. A method according to claim 3, wherein said second interest agent further comprises a fourth array comprising references to subscribing interest agents, and the second interest agent sends said information object to all interested agents in said fourth array if the user recommends said electronic information.

5. A method according to claim 4, wherein said second interest agent further comprises a fifth array of references to interest agents, and said second interest agent sends said information object to a specific number of random interest agents in said fifth array.

6. A method according to claim 5, wherein said second interest agent sends said information object to at least one interest agent in said third array, said fourth array or said fifth array.

7. A method according the claim 4, wherein if said second interest agent receives a message indicating a subscription from a third interest agent, then the reference of said third interest agent is added to said fourth array.

8. A method according to claim 1, further comprising:

first sorting the words in said electronic information in a second hash table giving values to said words in dependence on a) the number of occurrences, b) position in the electronic information, or c) tags, multiplying said second hash table with said first hash table thus receiving a number, wherein if said number is higher than a threshold value then the electronic information and said first model of interest match.

9. A method according to claim 8, wherein said second interest agent sends said information object to at least the interest agent in said third array which has the best match.

10. A method according to claim 8, wherein:

a user supplies an electronic document to an interface agent;

said interface agent creates a model of said electronic document;

said interface agent supplies said electronic document to an interest agent whose model of interest has the closest match to the model of said electronic information.

11. A method according to claim 1, wherein:

said second interest agent further comprises a sixth array comprising references to electronic information, and said information object further comprises a reference address to said electronic information;

when said second interest agent receives said information object from said first interest agent, the second interest agent adds the reference address to the electronic information to said sixth array if the confidence in said first interest agent is higher than a threshold value;

said second interest agent supplies said first model of interest and said sixth array to a search agent;

said search agent searches databases in the network for electronic information matching said first model of interest and avoids searching electronic information referenced in said sixth array; and

said search agent supplies electronic information to said second interest agent when a match occurs.

12. A method according to claim 1, wherein each position in said third array in said second interest agent further comprises a reference to a second model of interest for the interest agent that is referenced in the position.

13. A method according to claim 12, wherein said second model of interest comprises a third hash table, in which table a unique word has a unique position and where each position has a value indicating the importance of the corresponding word.

14. A method according to claim 12, wherein when said second interest agent receives an information object:

all second models of interest for corresponding interest agents referenced in the third array, which have an identical reference to any interest agent in said first array in said information object, are updated;

all references to interest agents in said first array which are not found in said third array are added to said third array; and

all references to interest agents in said second array which are not found in said fifth array are added to said fifth array.

15. A method according to claim 14, wherein:

said second model of interest comprises a third hash table, in which table a unique word has a unique position and where each position has a value indicating the importance of the corresponding word; and

said update is performed by determining a value for each unique word in said electronic information, wherein the value is determined in dependence on at least a) the number of occurrences of the unique word in said electronic information, b) the position in said electronic information of the unique word, or c) tags, and the determined value is added to the value at the position in said third hash table of the unique word.

16. A method according to claim 1, wherein said second interest agent sends a message to a third interest agent in said third array, indicating a subscription, if the confidence value corresponding to said third interest agent is higher than a threshold value.

17. A method according to claim 1, wherein:

said first interest agent adds its own reference to said first array in said information object if the user approves the present electronic information; and

said first interest agent adds its own reference to said second array in said information object if the user does not respond to the presentation of said electronic information.

18. A method according to claim 1, wherein:

if said electronic information is presented to a user and said user recommends said electronic information, then the confidence value in said first array for all interest agents which have recommended said electronic information is increased; and

if said electronic information is presented to a user and said user does not recommend said electronic information, then the confidence value in said first array for all interest agents which have recommended said electronic information is decreased.

19. A method according to claim 1, wherein the electronic information is presented to a user by displaying the electronic information on a monitor, and said user recommends said information by clicking on a button displayed on said monitor.