Machine assisted query formulation

• US 8,812,534 B2
• Filed: 05/25/2012
• Issued: 08/19/2014
• Est. Priority Date: 06/28/2007
• Status: Active Grant

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Platt, John C., Hurst-Hiller, Oliver, Gupta, Anoop, Flake, Gary W., Naam, Ramez, Goodman, Joshua T., Griffin, Trenholme J.
• Primary Examiner(s): KINSAUL, DANIEL W
• Application Number:
  US13/481,168
• Publication Number:
  US 20120239653A1
• Time in Patent Office: 816 Days
• US Class Current: 707/768
• CPC Class Codes: G06F 16/242 Query formulation

Abstract

Architecture for completing search queries by using artificial intelligence based schemes to infer search intentions of users. Partial queries are completed dynamically in real time. Additionally, search aliasing can also be employed. Custom tuning can be performed based on at least query inputs in the form of text, graffiti, images, handwriting, voice, audio, and video signals. Natural language processing occurs, along with handwriting recognition and slang recognition. The system includes a classifier that receives a partial query as input, accesses a query database based on contents of the query input, and infers an intended search goal from query information stored on the query database. A query formulation engine receives search information associated with the intended search goal and generates a completed formal query for execution.

Claims

1. A system comprising:

a processor;

a classifier to receive a partial query input for a search;

a syntactical processing component to analyze a structure of words or symbols used in the partial query input to generate content of the partial query input;

a database management system to index information to identify a database to access based at least in part on the content of the partial query input, the database including query information about search habits, search content, and recent used data of a user submitting the partial query input;

wherein the classifier, as a part of a machine learning and reasoning component, is further configured to infer a search goal of the user by retrieving from the database, similar or matching character sets, terms, or phrases, which are inferred over observations for completing the partial query input, the similar or the matching character sets, terms, or phrases being based at least in part on the indexed information and the query information; and

a query formulation component to generate a formal query based at least in part on the similar or the matching character sets, terms, or phrases and the partial query input;

wherein the machine learning and reasoning component is configured to receive an edit of the formal query from the user to determine a final query, analyze the edit of the formal query to determine a score indicative of a degree of success or failure of meeting the search goal of the user, and, if the score exceeds a threshold score, to store the final query as a learned response that is utilized to custom tune the query formulation component such that the query formulation component is configured to generate, based at least in part on the learned response, the final query in response to the classifier receiving, from the user, a subsequent partial query input that matches the partial query input.

2. The system of claim 1, further comprising a voice recognition component to receive and process the partial query input in a form of voice or speech signals.

3. The system of claim 1, further comprising a recognition component to receive and process the partial query input in a form of textual data, graffiti, handwriting, or short-hand.

4. The system of claim 1, further comprising an image component to receive and process the partial query input in a form of images or video frames.

5. The system of claim 1, further comprising an audio recognition component to receive and process the partial query input in a form of music.

6. The system of claim 1, wherein the machine learning and reasoning component is configured to employ a probabilistic or statistical-based analysis to prognose or infer a user action to be automatically performed.

7. The system of claim 1, further comprising a modeling component to generate a voice recognition model based on at least one of context data, location, terms, phrases, or speech characteristics.

8. A method executed on a processor for machine assisted query formulation, the method comprising:

receiving a partial query input for a search;

analyzing a structure of words or symbols used in the partial query input to generate content of the partial query input;

identifying a database to access based at least in part on the content of the partial query input, the database including query information about search habits, search content, and recent used data of a user submitting the partial query input;

inferring a search goal of the user by retrieving, from the database, similar or matching character sets, terms, or phrases, which are inferred over observations for completing the partial query input, the similar or the matching character sets, terms, or phrases being based at least in part on the query information;

generating a formal query based at least in part on the similar or the matching character sets, terms, or phrases and the partial query input;

receiving, from the user, an edit of the formal query to determine a final query;

analyzing the edit of the formal query to determine a score indicative of a degree of success or failure of meeting the search goal of the user;

storing the final query in the database as a learned response for formal query generation if the score exceeds a threshold score;

receiving, from the user, a subsequent partial query input that matches the partial query input; and

generating the final query based at least in part on the learned response.

9. The method of claim 8, further comprising receiving and processing the partial query input in a form of voice or speech signals.

10. The method of claim 8, further comprising receiving and processing the partial query input in a form of textual data, graffiti, handwriting, or short-hand.

11. The method of claim 8, further comprising receiving and processing the partial query input in a form of images or video frames.

12. The method of claim 8, further comprising receiving and processing the partial query input in a form of music.

13. The method of claim 8, further comprising employing a probabilistic or statistical-based analysis to prognose or infer a user action to be automatically performed.

14. The method of claim 8, further comprising generating a voice recognition model based on at least one of context data, location, terms, phrases, or speech characteristics.

15. A computer-readable storage device encoding instructions that when executed on a processor cause a computing device to perform acts comprising:

receiving a partial query input for a search;

analyzing a structure of words or symbols used in the partial query input to generate content of the partial query input;

identifying a database to access based at least in part on the content of the partial query input, the database including query information about search habits, search content, and recent used data of a user submitting the partial query input;

inferring a search goal of the user by retrieving, from the database, similar or matching character sets, terms, or phrases, which are inferred over observations for completing the partial query input, the similar or the matching character sets, terms, or phrases being based at least in part on the query information;

generating a formal query based at least in part on the similar or the matching character sets, terms, or phrases and the partial query input;

receiving, from the user, an edit of the formal query to determine a final query;

analyzing the edit of the formal query to determine a score indicative of a degree of success or failure of meeting the search goal of the user;

storing the final query in the database as a learned response for formal query generation if the score exceeds a threshold score;

receiving, from the user, a subsequent partial query input that matches the partial query input; and

generating the final query based at least in part on the learned response.

16. The computer-readable storage device of claim 15, wherein the acts further comprise receiving and processing the partial query input in a form of voice or speech signals.

17. The computer-readable storage device of claim 15, wherein the acts further comprise receiving and processing the partial query input in a form of images or video frames.

18. The computer-readable storage device of claim 15, wherein the acts further comprise employing a probabilistic or statistical-based analysis to prognose or infer a user action to be automatically performed.

19. The computer-readable storage device of claim 15, wherein the acts further comprise generating a voice recognition model based on at least one of context data, location, terms, phrases, or speech characteristics.