Determining confidence levels associated with attribute values of informational objects
Systems and methods for determining confidence levels associated with attribute values of informational objects. An example method comprises: receiving a natural language text; performing syntactico-semantic analysis of the natural language text to produce a plurality of semantic structures; interpreting the plurality of semantic structures using a set of production rules to produce a plurality of data items, each data item associating an attribute value with an informational object representing an entity referenced by the natural language text; and determining, for at least one data item of the plurality of data items, a confidence level associated with the attribute value, by evaluating a confidence function associated with the set of production rules.
1. A method, comprising:
training, using a training data set, a linear classifier producing a distance from an informational object to a hyperplane in a hyperspace of features, wherein the training data set comprises a first natural language text annotated to specify semantic classes associated with one or more words, and wherein accuracy of the classifier is increased after each training iteration;
receiving, by a processing device, a second natural language text;
performing syntactico-semantic analysis of the second natural language text to produce a plurality of semantic structures;
interpreting the plurality of semantic structures using a set of production rules to produce a plurality of attributes associated with a plurality of informational objects representing entities referenced by the second natural language text;
determining, for at least one informational object of the plurality of informational objects, a confidence level associated with the corresponding attribute value, by evaluating the linear classifier;
associating, by a data structure representing the second natural language text, the confidence level with the at least one informational object; and
utilizing the data structure for performing a natural language processing operation, wherein natural language processing operation includes at least one of: a machine translation operation, a semantic search operation, or a document classification operation.
2. The method of claim 1 , further comprising:
responsive to determining that the confidence level falls below a defined threshold, verifying the attribute value.
3. The method of claim 2 , wherein verifying the attribute value further comprises receiving, via a graphical user interface, a user input confirming the attribute value.
4. The method of claim 1 , further comprising:
producing a Resource Definition Framework (RDF) graph representing the natural language text.
5. The method of claim 1 , wherein each semantic structure of the plurality of semantic structures is represented by a graph comprising a plurality of nodes corresponding to a plurality of semantic classes and a plurality of edges corresponding to a plurality of semantic relationships.
6. The method of claim 1 , wherein a production rule of the set of production rules comprises one or more logical expressions defined on one or more semantic structure templates.
7. A system, comprising:
a memory;
a processor, coupled to the memory, the processor configured to:
train, using a training data set, a linear classifier producing a distance from an informational object to a hyperplane in a hyperspace of features, wherein the training data set comprises a first natural language text annotated to specify semantic classes associated with one or more words, and wherein accuracy of the classifier is increased after each training iteration;
receive a second natural language text;
perform syntactico-semantic analysis of the second natural language text to produce a plurality of semantic structures;
interpret the plurality of semantic structures using a set of production rules to produce a plurality of attributes associated with a plurality of informational objects representing entities referenced by the second natural language text;
determine, for at least one informational object of the plurality of informational objects, a confidence level associated with the corresponding attribute value, by evaluating the linear classifier; and
associate, by a data structure representing the second natural language text, the confidence level with the at least one informational object; and
utilize the data structure for performing a natural language processing operation wherein natural language processing operation includes at least one of: a machine translation operation, a semantic search operation, or a document classification operation.
8. The system of claim 7 , wherein the processor is further configured to:
responsive to determining that the confidence level falls below a defined threshold, verify the attribute value.
9. The system of claim 8 , wherein verifying the attribute value further comprises receiving, via a graphical user interface, a user input confirming the attribute value.
10. The system of claim 7 , wherein the processor is further configured to:
produce a Resource Definition Framework (RDF) graph representing the natural language text.
11. The system of claim 7 , wherein each semantic structure of the plurality of semantic structures is represented by a graph comprising a plurality of nodes corresponding to a plurality of semantic classes and a plurality of edges corresponding to a plurality of semantic relationships.
12. A computer-readable non-transitory storage medium comprising executable instructions that, when executed by a computer system, cause the computer system to:
train, using a training data set, a linear classifier producing a distance from an informational object to a hyperplane in a hyperspace of features, wherein the training data set comprises a first natural language text annotated to specify semantic classes associated with one or more words, and wherein accuracy of the classifier is increased after each training iteration;
receive a second natural language text;
perform syntactico-semantic analysis of the second natural language text to produce a plurality of semantic structures;
interpret the plurality of semantic structures using a set of production rules to produce a plurality of attributes associated with a plurality of informational objects representing entities referenced by the second natural language text;
determine, for at least one informational object of the plurality of informational objects, a confidence level associated with the corresponding attribute value, by evaluating the linear classifier; and
associate, by a data structure representing the second natural language text, the confidence level with the at least one informational object; and
utilize the data structure for performing a natural language processing operation, wherein natural language processing operation includes at least one of: a machine translation operation, a semantic search operation, or a document classification operation.
13. The computer-readable non-transitory storage medium of claim 12 , further comprising executable instructions causing the computer system to:
responsive to determining that the confidence level falls below a defined threshold, verify the attribute value.
14. The computer-readable non-transitory storage medium of claim 13 , wherein verifying the attribute value further comprises receiving, via a graphical user interface, a user input confirming the attribute value.