International Business Machines Corporation

Edward J Biddle

Avinash Asthana

Graham White

James Steven Luke

Stefan A G Van Der Stockt

Sujatha B Perepa

John Ferdinand Pitrelli

James Robert Magowan

Vijai Gandikota

Alex Swain

Daniel Kuehn

In an approach to creating explanatory confusion matrices, responsive to receiving a machine learning model for analysis, a confusion matrix is calculated for the machine learning model, where each cell in the confusion matrix has a corresponding set of data. A link is created from each cell in the confusion matrix to the corresponding set of data. Responsive to a user selecting in a user interface a specific cell of the confusion matrix, the corresponding set of data to the specific cell is displayed on the user interface.

Explanatory confusion matrices for machine learning

Embodiments provide a computer implemented method in a data processing system including a processor and a memory, the memory including instructions that are executed by the processor to cause the processor to implement a system for generating a type system. The method includes: receiving a document corpus; identifying frequently occurring words from the document corpus, disregarding stop words; extracting a conceptual text for each frequently occurring word from a structured information database; performing a cluster analysis on each conceptual text to identify possible entity types; performing a frequency analysis on possible entity types to select at least one entity type; identifying a relation type between entities in the document corpus; and generating the type system including entity types and relation types.

Generation of domain specific type system

Concepts for managing a supervised machine learning model of a set of documents are presented. A system obtains annotated versions of the documents, the documents being annotated by annotators. A conflict between a plurality of annotations of the annotated versions of the documents is identified. The machine learning model includes a set of entities and relations defining relationships between entities. The identified conflict is resolved by at least one of identifying the correct annotation between the conflicting options, splitting the annotated text into two separate entities or relations, generating a new entity at the same or a less specific hierarchical level as the entities or relation in conflict, and/or changing an annotation of the annotated version of the document.

Supervised machine learning models of documents

Systems, methods and computer program products for processing natural language input are provided. Natural language input is processed by one or more processing rules. The processing rules may specify one or more actions to be performed. The processing rules may alternatively or additionally split up the natural language input into a plurality of simpler sub-inputs, each of which may then be processed by one or more processing rules. The processing rules themselves may be generalized, which generalization may be based on user input. In the event that a suitable processing rule cannot be found, a request may be made to a user to provide instructions for processing the natural language input.

Natural language processing with adaptable rules based on user inputs

Growing community of inventors

Southampton, United Kingdom

Edward J Biddle