Extracting structured knowledge from unstructured text
AssigneeTrue Knowledge Ltd.
InventorsSimon Overell, William Tunstall-Pedoe
Life of this patent
Estimating the term…
Abstract
Embodiments of the present invention relate to knowledge representation systems which include a knowledge base in which knowledge is represented in a structured, machine-readable format that encodes meaning. Techniques for extracting structured knowledge from unstructured text and for determining the reliability of such extracted knowledge are also described.
Claims56
A computer-implemented method for extracting structured knowledge from unstructured text for use in a knowledge representation system, the knowledge representation system comprising a knowledge base that represents knowledge using a structured, machine-readable format, the structured, machine-readable format comprising fact triples, each fact triple including two entity objects and a relationship object expressing a relationship between the two entity objects, the method comprising:
- identifying sentences in the unstructured text using one or more computing devices;
- using the one or more computing devices, converting each of a subset of the sentences to one or more simplified assertion statements of the form:
- subject noun phrase, verb phrase, object noun phrase;
- converting each of a subset of the simplified assertion statements to a corresponding fact triple using the one or more computing devices;
- and
- determining a reliability for each of the fact triples using the one or more computing devices.
A system for extracting structured knowledge from unstructured text for use in a knowledge representation system, the knowledge representation system comprising a knowledge base that represents knowledge using a structured, machine-readable format, the structured, machine-readable format comprising fact triples, each fact triple including two entity objects and a relationship object expressing a relationship between the two entity objects, the system comprising one or more computing devices configured to:
- identify sentences in the unstructured text;
- convert each of a subset of the sentences to one or more simplified assertion statements of the form:
- subject noun phrase, verb phrase, object noun phrase;
- convert each of a subset of the simplified assertion statements to a corresponding fact triple;
- and
- determine a reliability for each of the fact triples.
A computer program product for extracting structured knowledge from unstructured text for use in a knowledge representation system, the knowledge representation system comprising a knowledge base that represents knowledge using a structured, machine-readable format, the structured, machine-readable format comprising fact triples, each fact triple including two entity objects and a relationship object expressing a relationship between the two entity objects, the computer program product comprising one or more non-transient computer-readable media having computer program instructions stored therein and configured, when executed, to cause one or more computing devices to:
- identify sentences in the unstructured text;
- convert each of a subset of the sentences to one or more simplified assertion statements of the form:
- subject noun phrase, verb phrase, object noun phrase;
- convert each of a subset of the simplified assertion statements to a corresponding fact triple;
- and
- determine a reliability for each of the fact triples.
A computer-implemented method for extracting structured knowledge from unstructured text for use in a knowledge representation system, the knowledge representation system comprising a knowledge base that represents knowledge using a structured, machine-readable format, the structured, machine-readable format comprising fact triples, each fact triple including two entity objects and a relationship object expressing a relationship between the two entity objects, the method comprising:
- identifying a question and a corresponding answer in the unstructured text using one or more computing devices;
- at least partially translating the question using the one or more computing devices;
- extracting additional information from the answer with reference to the partially translated question using the one or more computing devices;
- and
- using the one or more computing devices, generating one of the fact triples using the at least partially translated question and the additional information extracted from the answer.
A system for extracting structured knowledge from unstructured text for use in a knowledge representation system, the knowledge representation system comprising a knowledge base that represents knowledge using a structured, machine-readable format, the structured, machine-readable format comprising fact triples, each fact triple including two entity objects and a relationship object expressing a relationship between the two entity objects, the system comprising one or more computing devices configured to:
- identify a question and a corresponding answer in the unstructured text;
- at least partially translate the question;
- extract additional information from the answer with reference to the at least partially translated question;
- and
- generate one of the fact triples using the at least partially translated question and the additional information extracted from the answer.
A computer program product for extracting structured knowledge from unstructured text for use in a knowledge representation system, the knowledge representation system comprising a knowledge base that represents knowledge using a structured, machine-readable format, the structured, machine-readable format comprising fact triples, each fact triple including two entity objects and a relationship object expressing a relationship between the two entity objects, the computer program product comprising one or more non-transient computer-readable media having computer program instructions stored therein and configured, when executed, to cause one or more computing devices to:
- identify a question and a corresponding answer in the unstructured text;
- at least partially translate the question;
- extract additional information from the answer with reference to the at least partially translated question;
- and
- generate one of the fact triples using the at least partially translated question and the additional information extracted from the answer.
Dependencies are inferred from the claim text.