System and method for command line prediction

• US 20040117380A1
• Filed: 12/13/2002
• Published: 06/17/2004
• Est. Priority Date: 12/13/2002
• Status: Active Grant

• Current Assignee: Oracle America, Inc. (Oracle Corporation)
• Original Assignee: Sun Microsystems Incorporated (Oracle Corporation)
• Inventors: Perrow, Michael S.
• Granted Patent: US 7,284,009 B2
• Time in Patent Office: Days
• US Class Current: 707/100
• CPC Class Codes: G06F 3/0237 using prediction or retrieval techniques
• CPC Class Codes: G06F 9/45512 Command shells
• CPC Class Codes: Y10S 707/99936 Pattern matching access
• CPC Class Codes: Y10S 707/99943 Generating database or data structure, e.g. via user interface

Abstract

Systems, methods, and computer program products for command line prediction are disclosed. Commands entered on a command line are saved to a command log. The command log is analyzed to generate a plurality of n-gram tables that reflect the entries in the command log. The n-gram tables may be stored in a suitable memory. Subsequent command sequences entered on the command line may be compared to the n-gram tables to assess the likelihood of a command.

Claims

1. A method for command line prediction in a computer system including a database of command sequences organized into tables of n-grams, comprising:

receiving an entry in a command line sequence;

searching the database for an n-gram entry matching the received entry in the command line sequence; and

displaying a predicted command if an n-gram entry that satisfies a certainty threshold is found in the database.

2. The method of claim 1, wherein receiving an entry in a command line sequence comprises receiving a single command.

3. The method of claim 1, wherein receiving an entry in a command line sequence comprises storing the entry in a command line sequence in memory.

4. The method of claim 1, wherein searching the database for an n-gram entry matching the received entry in the command line sequence comprises searching for a 1-gram entry that matches the received entry.

5. The method of claim 4, wherein the command line sequence includes at least the two most recent commands entered at the command line and wherein matching 1-gram entries are used to search a 2-gram table for entries that match the two most recent commands entered at the command line.

6. The method of claim 5, wherein the command line sequence includes at least the three most recent commands entered at the command line and wherein matching 2-gram entries are used to search a 3-gram table for entries that match the two most recent commands entered at the command line.

7. The method of claim 6, wherein the command line sequence includes at least the four most recent commands entered at the command line and wherein matching 4-gram entries are used to search a 4-gram table for entries that match the two most recent commands entered at the command line.

8. The method of claim 7, wherein the command line sequence includes at least the five most recent commands entered at the command line and wherein matching 4-gram entries are used to search a 5-gram table for entries that match the two most recent commands entered at the command line.

9. The method of claim 1, further comprising determining a likelihood that the matching n-grams are predictive of the next command.

10. The method of claim 9, wherein the assigned likelihood is a function of the frequency of the matching n-grams and the total number of matching n-grams of a corresponding length.

11. A method for building a database for use in command line prediction, comprising:

storing commands entered by a user in a command log;

generating a plurality of n-gram tables from the commands stored in the command log, wherein the n-gram tables include an indicator of the frequency with which an n-gram appears in the command log; and

storing the n-gram tables in memory.

12. The method of claim 11, further comprising storing arguments associated with the commands in a command log.

13. The method of claim 11, wherein generating a plurality of n-gram tables comprises reading an entry from the command log into a data buffer and iteratively searching the n-gram tables for n-grams matching the entries in the data buffer.

14. A computer program product for command line prediction in a computer system including a database of command sequences organized into tables of n-grams, comprising:

logic instructions, executable on a processor, for receiving an entry in a command line sequence;

logic instructions, executable on a processor, for searching the database for an n-gram entry matching the received entry in the command line sequence; and

logic instructions, executable on a processor, for displaying a predicted command if an n-gram entry that satisfies a certainty threshold is found in the database.

15. The computer program product of claim 14, wherein logic instructions for receiving an entry in a command line sequence comprise logic instructions for receiving a single command.

16. The computer program product of claim 14, wherein logic instructions for receiving an entry in a command line sequence comprise logic instructions for storing the entry in a command line sequence in memory.

17. The computer program product of claim 14, wherein logic instructions for searching the database for an n-gram entry matching the received entry in the command line sequence comprise logic instructions for searching for a 1-gram entry that matches the received entry.

18. The computer program product of claim 17, wherein the command line sequence includes at least the two most recent commands entered at the command line and wherein matching 1-gram entries are used to search a 2-gram table for entries that match the two most recent commands entered at the command line.

19. The computer program product of claim 18, wherein the command line sequence includes at least the three most recent commands entered at the command line and wherein matching 2-gram entries are used to search a 3-gram table for entries that match the two most recent commands entered at the command line.

20. The computer program product of claim 19, wherein the command line sequence includes at least the four most recent commands entered at the command line and wherein matching 4-gram entries are used to search a 4-gram table for entries that match the two most recent commands entered at the command line.

21. The computer program product of claim 20, wherein the command line sequence includes at least the five most recent commands entered at the command line and wherein matching 4-gram entries are used to search a 5-gram table for entries that match the two most recent commands entered at the command line.

22. The computer program product of claim 14, further comprising determining logic instructions for a likelihood that the matching n-grams are predictive of the next command.

23. The computer program product of claim 22, wherein the assigned likelihood is a function of the frequency of the matching n-grams and the total number of matching n-grams of a corresponding length.

24. A system for command line prediction, comprising:

a processor configured to collect commands input by a user and store the commands in a command log;

a processor configured to retrieve entries from the command log and to generate a plurality of n-gram tables from the commands stored in the command log, wherein the n-gram tables include an indicator of the frequency with which an n-gram appears in the command log;

a database for storing the n-gram tables; and

a processor configured to receive an entry in a command line sequence input by a user, to search the database for an n-gram entry matching the received entry in the command line sequence; and

to display a predicted command if an n-gram entry that satisfies a certainty threshold is found in the database.

25. The system of claim 24, wherein the processor is configured to search the database for an n-gram entry matching the received entry in the command line sequence comprises searching for a 1-gram entry that matches the received entry.

26. The system of claim 25, wherein the command line sequence includes at least the two most recent commands entered at the command line and wherein matching 1-gram entries are used to search a 2-gram table for entries that match the two most recent commands entered at the command line.

27. The system of claim 26, wherein the command line sequence includes at least the three most recent commands entered at the command line and wherein matching 2-gram entries are used to search a 3-gram table for entries that match the two most recent commands entered at the command line.

28. The system of claim 27, wherein the command line sequence includes at least 115 the four most recent commands entered at the command line and wherein matching 4-gram entries are used to search a 4-gram table for entries that match the two most recent commands entered at the command line.

29. The system of claim 30, wherein the command line sequence includes at least the five most recent commands entered at the command line and wherein matching 4-gram entries are used to search a 5-gram table for entries that match the two most recent commands entered at the command line.

30. The system of claim 24, further comprising a processor configured to determine a likelihood that the matching n-grams are predictive of the next command.

31. The system of claim 30, wherein the assigned likelihood is a function of the frequency of the matching n-grams and the total number of matching n-grams of a corresponding length.