Providing natural-language interface to repository

1. A computer-program product tangibly embodied in a computer-readable storage device, the computer-program product comprising instructions that when executed cause a processor to perform operations for providing a natural-language interface to a repository, the operations comprising:

identifying at least one repository from which a new computer-readable ontology is to be created;

retrieving information from the at least one repository, wherein the retrieved information includes at least a first value of a first attribute and at least a second value of a second attribute;

determining a first data type of the first attribute based on the first value;

determining a second data type of the second attribute based on the second value;

generating the new computer-readable ontology using the retrieved information, the new computer-readable ontology indicating the first data type of the first attribute and the second data type of the second attribute; and

interpreting a user-entered natural-language statement for accessing the at least one repository using the new computer-readable ontology, including:

(a) identifying, in the first user-entered natural-language statement, a user-entered attribute name corresponding to the first attribute and the second attribute;

(b) identifying, in the first user-entered natural-language statement, a user-entered value associated with the user-entered attribute name;

(c) determining a data type of the user-entered value;

(d) comparing the data type of the user-entered value to the first data type of the first attribute and the second data type of the second attribute;

(e) selecting one of the first attribute and the second attribute based on the comparison; and

(f) using the selected attribute and the user-entered value in a predefined operation for accessing the at least one repository.

2. The computer-program product of claim 1 , wherein the information comprises metadata.

3. The computer-program product of claim 2 , wherein the metadata is retrieved from the repository, the metadata including a first table name representing a first table in the repository, at least a first attribute name representing the first attribute included in the first table, a second table name representing a second table in the repository, and at least a second attribute name representing the second attribute included in the second table.

4. The computer-program product of claim 3 , wherein several attributes are retrieved from the first table and wherein the computer-readable ontology upon being generated indicates an association between the several attributes.

5. The computer-program product of claim 1 , wherein several values associated with the first attribute are retrieved and several values associated with the second attribute are retrieved, and wherein the computer-readable ontology upon being generated indicates a first range of the several values of the first attribute and a second range of the several values of the second attribute.

6. The computer-program product of claim 1 , wherein the computer-readable ontology comprises a pointer structure associated with the repository.

7. The computer-program product of claim 1 , wherein the computer-readable ontology comprises an index of the repository.

8. The computer-program product of claim 1 , wherein interpretation comprises using at least one expression pattern.

9. A computer-program product tangibly embodied in a computer-readable storage device, the computer-program product comprising instructions that when executed cause a processor to perform operations for providing a natural-language interface to a repository, the operations comprising:

identifying at least one repository from which a new computer-readable ontology is to be created;

retrieving information from the at least one repository, wherein the retrieved information includes multiple first values of a first attribute and multiple second values of a second attribute;

determining a first range of the first values;

determining a second range of the second values;

generating the new computer-readable ontology using the retrieved information, the new computer-readable ontology indicating the first range of the first attribute and the second range of the second attribute;

interpreting a user-entered natural-language statement for accessing the at least one repository using the new computer-readable ontology, including:

(a) identifying, in the first user-entered natural-language statement, a user-entered attribute name corresponding to the first attribute and the second attribute;

(b) identifying, in the first user-entered natural-language statement, a user-entered value associated with the user-entered attribute name;

(c) comparing the user-entered value to the first range of the first attribute and the second range of the second attribute;

(d) selecting one of the first attribute and the second attribute based on the comparison; and

(e) using the selected attribute and the user-entered value in a predefined operation for accessing the at least one repository.

10. A computer system comprising:

at least one repository from which a new computer-readable ontology is to be created; and

an ontology-generating module generating the new computer-readable ontology using information retrieved from the at least one repository, the retrieved information including at least a first value of a first attribute and at least a second value of a second attribute, the new computer-readable ontology indicating a first data type of the first attribute based on the first value and a second data type of the second attribute based on the second value;

an input device that receives a first user-entered natural-language statement for accessing the at least one repository; and

an interpreting module that uses the computer-readable ontology in interpreting the first user-entered natural-language statement, wherein the interpreting module:

(a) identifies, in the first user-entered natural-language statement, a user-entered attribute name corresponding to the first attribute and the second attribute;

(b) identifies, in the first user-entered natural-language statement, a user-entered value associated with the user-entered attribute name;

(c) determines a data type of the user-entered value;

(d) compares the data type of the user-entered value to the first data type of the first attribute and the second data type of the second attribute;

(e) selects one of the first attribute and the second attribute based on the comparison; and

(f) uses the selected attribute and the user-entered value in a predefined operation for accessing the at least one repository.

11. The computer system of claim 10 , wherein the ontology-generating module retrieves metadata from the repository, the metadata including a first table name representing a first table in the repository, at least a first attribute name representing the first attribute included in the first table, a second table name representing a second table in the repository, and at least a second attribute name representing the second attribute included in the second table.

12. The computer system of claim 10 , wherein the ontology-generating module retrieves several values from the repository associated with the first attribute and several values associated with the second attribute, and generates the computer-readable ontology to indicate a first range of the several values of the first attribute and a second range of the several values of the second attribute.