Method and apparatus for automated, context-dependent retrieval of information

• US 6,236,768 B1
• Filed: 05/01/1998
• Issued: 05/22/2001
• Est. Priority Date: 10/14/1997
• Status: Active Grant

• Current Assignee: Massachusetts Institute of Technology
• Original Assignee: Massachusetts Institute of Technology
• Inventors: Rhodes, Bradley J., Pentland, Alex P., Starner, Thad E., Maes, Pattie E.
• Primary Examiner(s): Couso, Jose L.
• Assistant Examiner(s): DESIRE, GREGORY M
• Application Number:
  US09/071,312
• Time in Patent Office: 1,117 Days
• Field of Search: 382/305-306, 382/105, 382/155-159, 382/224-229, 707/1-3, 707/6, 707/200
• US Class Current: 382/306
• CPC Class Codes: G06F 16/38 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: G06F 16/907 Retrieval characterised by using metadata, e.g. metadata not derived from the content or metadata generated manually
• CPC Class Codes: Y10S 707/99931 Database or file accessing

Abstract

Documents stored in a database are searched for relevance to contextual information, instead of (or in addition to) similar text. Each stored document is indexed in term of meta-information specifying contextual information about the document. Current contextual information is acquired, either from the user or the current computational or physical environment, and this “meta-information” is used as the basis for identifying stored documents of possible relevance.

Claims

1. Apparatus for context-based document identification, the apparatus comprising:

a. a database indexing a plurality of documents each in terms of meta-information specifying contextual information about the document;

b. means for automatically acquiring current contextual information;

c. a search facility for searching the database to identify documents whose meta-information comprises information relevant to the acquired current contextual information; and

d. a display for reporting the identified documents to the user.

2. The apparatus of claim 1 wherein the meta-information comprises at least one of (a) a user location, (b) time of day, (c) day of week, (d) date, and (e) subject.

3. The apparatus of claim 1 wherein the meta-information comprises identification of a person associated with the document.

4. The apparatus of claim 1 wherein the means for acquiring current contextual information comprises an environmental sensor.

5. The apparatus of claim 4 wherein the environmental sensor is a global-positioning system.

6. The apparatus of claim 1 wherein the means for acquiring current contextual information comprises means for identifying a nearby individual.

7. The apparatus of claim 1 further comprising a system clock, the means for acquiring current contextual information being connected to the system clock and deriving contextual information therefrom.

8. The apparatus of claim 1 wherein at least some of the meta-information is continuous, the searching means automatically and continuously identifying relevant information based on proximity of the meta-information to the current contextual information.

9. The apparatus of claim 1 wherein at least some of the meta-information is discrete, the searching means identifying relevant information based on an exact match between the meta-information and the current contextual information.

10. The apparatus of claim 1 further comprising means for adding new documents to the database, said means comprising:

a. a plurality of document templates, each template corresponding to a document type and specifying contextual information within the document; and

b. an analyzer for (i) matching a new document to a template and, in accordance with the template, extracting contextual information from the document, and (ii) indexing the document within the database in terms of the extracted contextual information.

11. The apparatus of claim 1 wherein the meta-information is represented in the database as vectors, each vector corresponding to a document and to a type of contextual information associated therewith and having a value representative of the associated contextual information.

12. The apparatus of claim 11 wherein the current contextual information is represented as a vector, the searching means determining relevance based on the current contextual-information vector and the vectors in the database.

13. The apparatus of claim 1 wherein the means for acquiring current contextual information comprises user-responsive means for accepting user-provided contextual information.

14. The apparatus of claim 1 further comprising means for storing a current document, the means for acquiring current contextual information comprising means for analyzing the current document for contextual information and extracting the contextual information therefrom.

15. A method of identifying documents from a stored document database in response to contextual information, the method comprising the steps of:

a. indexing each stored document in terms of meta-information specifying contextual information about the document;

b. automatically acquiring current contextual information;

c. identifying stored documents whose meta-information comprises information relevant to the acquired current contextual information; and

d. reporting the located documents to the user.

16. The method of claim 15 wherein the meta-information comprises at least one of (a) a user location, (b) time of day, (c) day of week, (d) date, and (e) subject.

17. The method of claim 15 wherein the meta-information comprises identification of a person associated with the document.

18. The method of claim 15 wherein at least some of the meta-information is continuous, relevant information being automatically and continuously identified based on proximity of the meta-information to the current contextual information.

19. The method of claim 15 wherein at least some of the meta-information is discrete, relevant information being identified based on an exact match between the meta-information and the current contextual information.

20. The method of claim 15 further comprising the step of adding new documents to the database according to substeps comprising:

a. defining a plurality of document templates, each template corresponding to a document type and specifying contextual information within the document;

b. matching a new document to a template and, in accordance with the template, extracting contextual information from the document; and

c. indexing the document within the database in terms of the extracted contextual information.

21. The method of claim 15 further comprising the step of representing the meta-information as vectors, each vector corresponding to a document and to a type of contextual information associated therewith and having a value representative of the associated contextual information.

22. The method of claim 21 wherein the current contextual information is represents as a vector, relevance being determined in accordance with the current contextual information vector and the vectors in the database.

23. The method of claim 15 wherein the current contextual information is acquired from a user.

24. The method of claim 15 wherein current contextual information is acquired from a current document.