INFORMATION EXTRACTION FROM SEMANTIC DATA
Technologies and implementations for extracting information from semantic data available, for example, on the World Wide Web, are generally disclosed.
1 . A method to extract information from semantic data on the world wide web, the method comprising:
generating a plurality of assertions from an ontology corresponding to the semantic data based at least in part on a plurality of statements of the ontology;
determining information candidates based at least in part on syntax of information representation language; and
validating the information candidates based at least in part on the plurality of assertions.
2 . The method of claim 1 , wherein generating the plurality of assertions comprises generating one or more assertions based at least in part upon a terminological box (Tbox) classification and an assertion box (Abox) sampling.
3 . The method of claim 2 , wherein generating the plurality of assertions comprises determining a concept hierarchy tree and a role hierarchy tree, both being based at least in part on the Tbox classification.
4 . The method of claim 2 , wherein generating the plurality of assertions comprises determining an assertion pattern based at least in part on the Abox sampling.
5 . The method of claim 4 , wherein determining the assertion pattern comprises generating a plurality of distilled assertions based at least in part on the Abox sampling and the Tbox classification.
6 . The method of claim 1 , wherein determining information candidates comprises determining information candidates based at least in part on a description logic.
7 . The method of claim 6 , wherein determining information candidates based at least in part on the description logic comprises determining information candidates based at least in part on Web Ontology Language (OWL).
8 . The method of claim 1 , wherein determining information candidates comprises determining information candidates based at least in part on syntax of information representation language and signatures included in the Tbox classification.
9 . The method of claim 1 , wherein determining information candidates comprises determining information candidates based at least in part on novelty rule.
10 . The method of claim 1 , wherein determining information candidates comprises determining information candidates based at least in part on simplicity rule.
11 . The method of claim 1 , wherein validating the information candidates comprises determining an approximate Abox sampling.
12 . The method of claim 1 , wherein validating the information candidates comprises calculating a certainty level for a concept candidate based at least in part on a majority rule.
13 . A machine readable non-transitory medium having stored therein instructions that, when executed by one or more processors, operatively enable a semantic data processing module to:
generate a plurality of assertions from an ontology corresponding to the semantic data based at least in part on a terminological box (Tbox) classification and an assertion box (Abox) sampling;
determine information candidates based at least in part on syntax of information representation language; and
validate the information candidates based at least in part on the plurality of assertions.
14 . The machine readable non-transitory medium of claim 13 , wherein the stored instructions, when executed by one or more processors, further operatively enable the semantic data processing module to determine a concept hierarchy tree and a role hierarchy tree, both being based at least in part on the Tbox classification.
15 . The machine readable non-transitory medium of claim 14 , wherein the stored instructions, when executed by one or more processors, further operatively enable the semantic data processing module to assign instances to at least one of concepts and roles based at least in part on the concept hierarchy tree and the role hierarchy tree.
16 . The machine readable non-transitory medium of claim 13 , wherein the stored instructions, when executed by one or more processors, further operatively enable the semantic data processing module to determine an assertion pattern based at least in part on the Abox sampling.
17 . The machine readable non-transitory medium of claim 16 , wherein the stored instructions, when executed by one or more processors, further operatively enable the semantic data processing module to generate a plurality of distilled assertions based at least in part on the Abox sampling and the Tbox classification.
18 . The machine readable non-transitory medium of claim 13 , wherein the stored instructions, when executed by one or more processors, further operatively enable the semantic data processing module to determine information candidates based at least in part on a description logic.
19 . The machine readable non-transitory medium of claim 18 , wherein the stored instructions, when executed by one or more processors, further operatively enable the semantic data processing module to determine information candidates based at least in part on Web Ontology Language (OWL).
20 . The machine readable non-transitory medium of claim 13 , wherein the stored instructions, when executed by one or more processors, further operatively enable the semantic data processing module to determine information candidates based at least in part on syntax of information representation language and signatures included in the Tbox classification.
21 . The machine readable non-transitory medium of claim 13 , wherein the stored instructions, when executed by one or more processors, further operatively enable the semantic data processing module to determine an approximate Abox sampling.
22 . The machine readable non-transitory medium of claim 13 , wherein the stored instructions, when executed by one or more processors, further operatively enable the semantic data processing module to calculate a certainty level for a concept candidate based at least in part on a majority rule.
23 . A system to extract information from semantic data on the world wide web comprising:
a processor; and
a semantic data processing module communicatively coupled to the processor, the semantic data processing module configured to:
generate a plurality of assertions from an ontology corresponding to the semantic data based at least in part on a terminological box (Tbox) classification and an assertion box (Abox) sampling;
determine information candidates based at least in part on syntax of information representation language; and
validate the information candidates based at least in part on the plurality of assertions.
24 . The system of claim 23 , wherein the semantic data processing module is further configured to determine a concept hierarchy tree and a role hierarchy tree, both being based at least in part on the Tbox classification.
25 . The system of claim 24 , wherein the semantic data processing module is further configured to assign instances to at least one of concepts and roles based at least in part on the concept hierarchy tree and the role hierarchy tree.
26 . The system of claim 23 , wherein the semantic data processing module is further configured to determine an assertion pattern based at least in part on the Abox sampling.
27 . The system of claim 26 , wherein the semantic data processing module is further configured to generate a plurality of distilled assertions based at least in part on the Abox sampling and the Tbox classification.
28 . The system of claim 23 , wherein the semantic data processing module is further configured to determine information candidates based at least in part on a description logic.
29 . The system of claim 28 , wherein the semantic data processing module is further configured to determine information candidates based at least in part on Web Ontology Language (OWL).
30 . The system of claim 23 , wherein the semantic data processing module is further configured to determine information candidates based at least in part on syntax of information representation language and signatures included in the Tbox classification.
31 . The system of claim 23 , wherein the semantic data processing module is further configured to determine an approximate Abox sampling.
32 . The system of claim 23 , wherein the semantic data processing module is further configured to calculate a certainty level for a concept candidate based at least in part on a majority rule.