Electronic assistant for making predictions based on user messages

• US 9,443,226 B2
• Filed: 08/10/2009
• Issued: 09/13/2016
• Est. Priority Date: 01/12/2009
• Status: Active Grant

• Current Assignee: SRI International, Inc.
• Original Assignee: SRI International, Inc.
• Inventors: Berry, Pauline M., Cheyer, Adam, Gervasio, Melinda, Donneau-Golencer, Thierry, Hardt, Stephen, Nitz, Kenneth, Peintner, Bart, Pound, Leslie, Yorke-Smith, Neil
• Primary Examiner(s): Hoang, Son T
• Application Number:
  US12/538,475
• Publication Number:
  US 20100179961A1
• Time in Patent Office: 2,591 Days
• Field of Search: 707/802, 707/769, 706/47, 709/206
• US Class Current: 1/1
• CPC Class Codes: G06F 16/80 of semi-structured data, e.g. markup language structured data such as SGML, XML or HTML content-based retrieval of web data G06F16/95
• CPC Class Codes: G06F 40/284 Lexical analysis, e.g. tokenisation or collocates
• CPC Class Codes: G06Q 10/107 Computer-aided management of electronic mailing [e-mailing]

Abstract

The present invention relates to an electronic assistant. In one embodiment, a system for processing data on a computing device includes at least one extractor for extracting data from an email to produce extracted data, a user model for generating a high-level model of an email environment on the computing device, based on the extracted data, and at least one predictor for generating a prediction related to the email, based on the extracted data and on the high-level model. In another embodiment, a system for scheduling an event includes an extractor for extracting constraints from a request for the event, a user model for receiving and storing scheduling preferences from a user, and a predictor coupled to the extractor and to the user model, for generating a candidate schedule for the event in accordance with the constraints and the scheduling preferences.

Claims

1. A natural language tagging system for processing electronic files on a user'"'"'s computing system, wherein the natural language tagging system is embodied in one or more non-transitory computer accessible storage media and comprises instructions executable by a processor to:

observe tagging of an electronic file by a user on the computing system, the user tagging producing a user-tagged electronic file;

associate a user-created natural language query with a tag of the user-tagged electronic file, wherein the user-created natural language query is received via a user interface of the computing system and relates to a plurality of electronic files stored on the computing system;

execute the natural language query on the plurality of electronic files stored on the computing system;

compute a similarity between each of one or more untagged electronic files included in a search result for the executed natural language query;

in an automated fashion, tag an untagged electronic file of the one or more untagged electronic files in the search result with a tag comprising at least a portion of the natural language query;

wherein the natural language tagging system determines a folder name of a folder storing each of the one or more untagged electronic files on the computing system,

wherein, in an automated fashion, the natural language tagging system propagates at least a portion of the folder name as a part of the tag for each of the one or more untagged electronic files stored in the folder, and

wherein the system determines a frequent word contained in each of the one or more untagged electronic files stored in the folder and propagates the folder name as the part of the tag for the one or more untagged electronic files stored in the folder only if the folder name matches the frequent word.

2. A natural language tagging system for processing electronic files on a user'"'"'s computing system, wherein the natural language tagging system is embodied in one or more non-transitory computer accessible storage media and comprises instructions executable by a processor to:

observe tagging of an electronic file by a user on the computing system, the user tagging producing a user-tagged electronic file;

associate a tag of the user-tagged electronic file with a user-created natural language query, wherein the user-created natural language query is received via a user interface of the computing system, the natural language query relating to a plurality of electronic files stored on the computing system;

execute the natural language query on the plurality of electronic files stored on the computing system;

compute a similarity between each of one or more untagged electronic files included in a search result for the executed natural language query;

in an automated fashion, tag an untagged electronic file of the one or more untagged electronic files in the search result with a tag that comprises at least a portion of the natural language query;

determine a frequent word contained in the particular untagged electronic file stored on the computing system,

execute an external service using the frequent word as input to the external service, and

suggest, as a part of the tag, for the particular untagged electronic file based on information returned by the external service in response to the executing of the external service with the frequent word as input.

3. An electronic assistant being embodied in one or more non-transitory computer accessible storage media and comprising instructions executable by a processor to tag a plurality of electronic files stored on a user'"'"'s computing system by:

observing tagging of an electronic file by the user on the computing system, the user tagging producing a user-tagged electronic file;

associating a tag of the user-tagged electronic file with a user-created natural language query;

computing a similarity between an untagged electronic file on the computing system and the user-tagged electronic file;

in an automated fashion, tagging the untagged electronic file with the tag of the user-tagged electronic file in response to a determination that the computed similarity meets or exceeds a predefined threshold;

harvesting the plurality of electronic files stored on the computing system;

creating a dictionary of most important words contained in the plurality of harvested electronic files;

receiving a new electronic file, wherein the new electronic file comprises the tag of the user-tagged electronic file;

creating a bag of words model representing content of the new electronic file;

computing a similarity score indicating a level of similarity of at least one word in the bag of words model to at least one word in the dictionary; and

creating a new tag for the new electronic file, the new tag comprising a particular word of the bag of words model in response to the similarity score indicating a high level of similarity between the particular word of the bag of words model and the at least one word in the dictionary;

periodically re-executin the computing of the similarity score and deleting the created tag of the new electronic file in response to the similarity score failing to indicate a high level of similarity between the at least one word in the bag of words model and the at least one word in the dictionary.