Corpus gap probability modeling
AssigneeInternational Business Machines Corporation
InventorsMichael McCawley, Adriana Valido, Robert E. Loredo
Life of this patent
Estimating the term…
Abstract
A system, computer program product, and method are provided to conduct gap probability mapping to predict presence and location of one or more gaps in a corpus. A probabilistic model is formed to represent inter-corpora associations of objects, which the model leverages to process query submissions. As queries are received and processed, the model creates an adjustment. Subject to evaluation, confidence of the adjustment is evaluated, and a response correlated to the confidence is returned.
Claims15
A system comprising:
- a processing unit operatively coupled to memory;
- an artificial intelligence platform in communication with the processing unit and memory;
- and
- a knowledge engine operatively coupled to the processing unit, the knowledge engine to assess probability and generate a map based on the assessment, including:
- an organization manager configured to:
- for corpora having respective pluralities of objects, organize the objects into a plurality of taxonomies, including organize the respective plurality of objects of each corpus of the corpora into an associated taxonomy based on object characteristic data;
- and
- analyze the corpora according to the associated taxonomies, the analysis including formation of an inter-corpora ontology comprising one or more inter-corpora associations of the respective objects;
- a query manager operatively coupled to the organization manager, the query manager configured to:
- receive and process a query and associated query content, including leverage the inter-corpora ontology to identify, from the plurality of taxonomies, a taxonomy associated with the query content;
- submit the query to the corpus associated with the identified taxonomy;
- receive an initial response with an initial probability proportional to a frequency of the initial response in the corpus associated with the identified taxonomy;
- and
- apply the query content to the inter-corpora ontology, the application to form an adjusted probability and an adjusted response;
- and
- a director operatively coupled to the query manager, the director configured to assess confidence of accuracy of a response to the query based on evaluation of the initial probability value and the adjusted probability value;
- and
- the query manager to return a response, the response directly correlated to the assessed confidence conducted by the director, wherein the response includes the initial response or the adjusted response.
A computer program product comprising:
- a computer readable storage medium having computer readable program code embodied therewith, the program code being executable by a processor to assess probability and generate a map based on the assessment, including program code executable by the processor to:
- for corpora having respective pluralities of objects, organize the objects into a plurality of taxonomies, including organize the respective plurality of objects of each corpus of the corpora into an associated taxonomy based on object characteristic data;
- analyze the corpora according to the associated taxonomies, the analysis including formation of an inter-corpora ontology comprising one or more inter-corpora associations of the respective objects;
- receive and process a query and associated query content, including leverage the inter-corpora ontology to identify, from the plurality of taxonomies, a taxonomy associated with the query content;
- submit the query to the corpus associated with the identified taxonomy;
- receive an initial response with an initial probability proportional to a frequency of the initial response in the corpus associated with the identified taxonomy;
- apply the query content to the inter-corpora ontology, the application to form an adjusted probability and an adjusted response;
- assess confidence of accuracy of a response to the query based on evaluation of the initial probability value and the adjusted probability value;
- and
- return a response, the response directly correlated to the assessed confidence, wherein the response includes the initial response or the adjusted response.
A method comprising:
- for corpora having respective pluralities of objects, organizing the objects into a plurality of taxonomies, including organize the respective plurality of objects of each corpus of the corpora into an associated taxonomy based on object characteristic data;
- analyzing the corpora according to the associated taxonomies, the analysis including forming an inter-corpora ontology comprising one or more inter-corpora associations of the respective objects;
- receiving and processing a query and associated query content, including leveraging the inter-corpora ontology to identify, from the plurality of taxonomies, a taxonomy associated with the query content;
- submitting the query to the corpus associated with the identified taxonomy;
- receiving an initial response with an initial probability proportional to a frequency of the initial response in the corpus associated with the identified taxonomy;
- applying the query content to the inter-corpora ontology, the application to form an adjusted probability and an adjusted response;
- assessing confidence of accuracy of a response to the query based on evaluation of the initial probability value and the adjusted probability value;
- and
- returning a response from the confidence assessment, the response directly correlated to the assessed confidence, wherein the response includes the initial response or the adjusted response.
Dependencies are inferred from the claim text.