Lexical, grammatical, and semantic inference mechanisms

• US 20060277525A1
• Filed: 06/06/2005
• Published: 12/07/2006
• Est. Priority Date: 06/06/2005
• Status: Abandoned Application

• Current Assignee: Microsoft Technology Licensing LLC (Microsoft Corporation)
• Original Assignee: Microsoft Corporation
• Inventors: Horton, Anson M., Najmabadi, Cyrus
• Application Number:
  US11/147,082
• Publication Number:
  US 20060277525A1
• Time in Patent Office: Days
• US Class Current: 717/106
• CPC Class Codes: G06F 8/33 Intelligent editors
• CPC Class Codes: G06N 5/04 Inference or reasoning models

Abstract

A mechanism is provided for inferring a computer user'"'"'s intent in developing computer system code. Specifically, an inference mechanism is provided that can infer a user'"'"'s intent at least on three levels: a lexical level, a syntactic level, and a semantic level. The inference mechanism employs various rule-based and heuristic-based techniques including type coercion, scope proximity, parameter count and arguments, and so on. Various inference mechanism controls are also employed to increase inference robustness, including timing of making suggestions, the number of suggestions, and the extensibility of suggestions.

Claims

1. A method for inferring a user'"'"'s intent in developing a computer code, comprising:

obtaining at least one token of the computer code developed by the user;

using an inference mechanism, wherein the inference mechanism examines the at least one token of the computer code to find out if a mistake has been made by the user in developing the at least one token, wherein the mistake is determined by examining at least one of a lexical rule and a syntactic rule, and at least one of a semantic heuristic; and

providing at least one of a suggestion and a correction to remedy the mistake, wherein the inference mechanism provides the at least one suggestion and correction according to a control standard.

2. The method according to claim 1, wherein examining the at least one of the lexical rule and the syntactic rule, and the at least one of the semantic heuristic includes comparing the scope proximity of the at least one token to at least another token in the code.

3. The method according to claim 1, wherein examining the at least one of the lexical rule and the syntactic rule, and the at least one of the semantic heuristic includes comparing variable types in the code.

4. The method according to claim 1, wherein examining the at least one of the lexical rule and the syntactic rule, and the at least one of the semantic heuristic includes examining the number of parameters and arguments in functions and function calls, respectively.

5. The method according to claim 1, wherein examining the at least one of the lexical rule and the syntactic rule, and the at least one of the semantic heuristic is based on feedback of how the user is developing the code.

6. The method according to claim 1, wherein providing the at least one of the suggestion and the correction to remedy the mistake is based on a percentage of confidence that the inference mechanism is correct.

7. The method according to claim 1, wherein the inference mechanism is extensible so that additional lexical and syntactic rules and semantic heuristics can be added to the at least one lexical and syntactic rules and the at least one semantic heuristic in order to improve inferring the user'"'"'s intent.

8. A computer readable medium bearing computer readable instruction for inferring a user'"'"'s intent in developing a computer code, comprising:

receiving a plurality of tokens of the computer code developed by a developer;

subjecting the plurality of tokens to an inference mechanism, wherein the inference mechanism examines the plurality of tokens to find out if an error has been committed by the developer in developing the tokens, wherein the error is determined by examining one of a lexical rule and a syntactic rule, and one of a semantic heuristic; and

providing at least one of a suggestion and a correction to remedy the error, wherein the inference mechanism provides the at least one suggestion and correction according to a predetermined standard.

9. The computer readable medium according to claim 8, wherein examining one of the lexical rule and the syntactic rule, and the semantic heuristic includes comparing the scope proximity of at least one token in the plurality of tokens to at least one other token in the plurality of tokens.

10. The computer readable medium according to claim 8, wherein examining one of the lexical rule and the syntactic rule, and the semantic heuristic includes comparing variable types in the code.

11. The computer readable medium according to claim 8, wherein examining one of the lexical rule and the syntactic rule, and the semantic heuristic includes examining the number of parameters and arguments in functions and function calls, respectively.

12. The computer readable medium according to claim 8, wherein examining one of the lexical rule and the syntactic rule, and the semantic heuristic is based on feedback of how the developer is developing the code.

13. The computer readable medium according to claim 8, wherein providing at least one of the suggestion and the correction to remedy the error is based on a percentage of confidence that the inference mechanism is correct.

14. The computer readable medium according to claim 8, wherein the inference mechanism is extensible so that additional lexical and syntactic rules and semantic heuristics can be added to the inference mechanism.

15. An apparatus for inferring a user'"'"'s actions in developing a computer code, comprising:

an inference mechanism, wherein the inference mechanism examines token by token any input of tokens in order to determine any mistakes by the user in developing the tokens, wherein any mistakes are determined according to predetermined rules and encoded heuristics, and wherein the inference mechanism provides one of a suggestion and a correction upon determining of any of the mistakes according to a control standard.

16. The apparatus according to claim 15, wherein examining the predetermined rules and encoded heuristics include comparing the scope proximity of the tokens.

17. The apparatus according to claim 15, wherein examining the predetermined rules and encoded heuristics includes comparing variable types in the code.

18. The apparatus according to claim 15, wherein examining the predetermined rules and encoded heuristics includes examining the number of parameters and arguments in functions and function calls, respectively.

19. The apparatus according to claim 15, wherein examining the predetermined rules and encoded heuristics is based on feedback of how the user is developing the code.

20. The apparatus according to claim 15, wherein the inference mechanism is extensible so that additional predetermined rules and encoded heuristics can be added to the inference mechanism.