Information extraction from natural language texts
Systems and methods for extracting facts from natural language texts. An example method of information extraction comprises extracting, from a natural language text, a first plurality of information objects; extracting, from the natural language text, a second plurality of information objects; identifying a set of conflicting information objects, such that a first information object of the set of conflicting information objects belongs to the first plurality of information objects and a second information object of the set of conflicting information objects belongs to the second plurality of information objects; and producing a final list of information objects extracted from the natural language text, by applying, to the set of conflicting information objects, a conflict arbitration function which performs at least one of: modifying the first information object, deleting the first information object, or merging two or more information objects of the set of conflicting information objects.
1. A method, comprising:
extracting, by a computer system, a first plurality of information objects from a natural language text;
extracting, from the natural language text, a second plurality of information objects;
determining that a first textual annotation associated with a first information object of the first plurality of information objects is overlapping with a second textual annotation associated with a second information object of the second plurality of information objects;
applying, to the first information object and the second information object, a conflict arbitration function represented by a machine learning classifier yielding a likelihood of the first information object and the second information object representing a same object.
2. The method of claim 1 , wherein extracting the first plurality of information objects is performed by a first information extraction technique and extracting the second plurality of information objects is performed by a second information extraction technique.
3. The method of claim 1 , further comprising:
producing a final list of information objects extracted from the natural language text; and
utilizing the final list of information objects for performing a natural language processing operation.
4. The method of claim 1 , further comprising:
producing a final list of information objects extracted from the natural language text; and
representing the final list of information objects by a Resource Definition Framework (RDF) graph.
5. The method of claim 1 , further comprising:
evaluating a logical condition comprising a first attribute of the first information object and a second attribute of the second information object.
6. The method of claim 1 , further comprising:
determining that the first information object has a number of attributes of a certain type exceeding a threshold number of attributes of the certain type.
7. The method of claim 1 , further comprising:
appending, to a training data set, the natural language text accompanied by metadata comprising definitions and textual annotations of the first information object and the second information object; and
training, utilizing the training data set, a machine learning classifier implementing the conflict arbitration function.
8. The method of claim 1 , further comprising: determining a first confidence level associated with the first information object.
9. A computer system, comprising:
a memory;
a processor, coupled to the memory, the processor configured to:
extract a first plurality of information objects from a natural language text;
extract, from the natural language text, a second plurality of information objects;
determine that a first textual annotation associated with a first information object of the first plurality of information objects is overlapping with a second textual annotation associated with a second information object of the second plurality of information objects; and
apply, to the first information object and the second information object, a conflict arbitration function represented by a machine learning classifier yielding a likelihood of the first information object and the second information object representing a same object.
10. The computer system of claim 9 , wherein extracting the first plurality of information objects is performed by a first information extraction technique and extracting the second plurality of information objects is performed by a second information extraction technique.
11. The computer system of claim 9 , wherein the processor is further configured to:
produce a final list of information objects extracted from the natural language text; and
utilize the final list of information objects for performing a natural language processing operation.
12. The computer system of claim 9 , wherein the processor is further configured to:
produce a final list of information objects extracted from the natural language text; and
represent the final list of information objects by a Resource Definition Framework (RDF) graph.
13. The computer system of claim 9 , wherein the processor is further configured to:
evaluate a logical condition comprising a first attribute of the first information object and a second attribute of the second information object.
14. The computer system of claim 9 , wherein the processor is further configured to:
append, to a training data set, the natural language text accompanied by metadata comprising definitions and textual annotations of the first information object and the second information object; and
train, utilizing the training data set, a machine learning classifier implementing the conflict arbitration function.
15. The computer system of claim 9 , wherein the processor is further configured to: determine a first confidence level associated with the first information object.
16. A computer-readable non-transitory storage medium comprising executable instructions that, when executed by a computer system, cause the computer system to:
extract a first plurality of information objects from a natural language text;
extract, from the natural language text, a second plurality of information objects;
determine that a first textual annotation associated with a first information object of the first plurality of information objects is overlapping with a second textual annotation associated with a second information object of the second plurality of information objects; and
apply, to the first information object and the second information object, a conflict arbitration function represented by a machine learning classifier yielding a likelihood of the first information object and the second information object representing a same object.
17. The computer-readable non-transitory storage medium of claim 16 , wherein extracting the first plurality of information objects is performed by a first information extraction technique and extracting the second plurality of information objects is performed by a second information extraction technique.
18. The computer-readable non-transitory storage medium of claim 16 , further comprising executable instructions that, when executed by the computer system, cause the computer system to:
produce a final list of information objects extracted from the natural language text; and
utilize the final list of information objects for performing a natural language processing operation.