International Business Machines Corporation

Lianhua Chi

Christopher John Butler

Bo Han

Alessio Bonti

Juerg Von Kaenel

Anna T T Phan

A method for identifying correlated datasets comprises receiving a natural language query from a user, extracting electronic metadata from a plurality of computerized datasets, correlating the electronic metadata with the natural language query using natural language processing, calculating a correlation score between the natural language query and each of the plurality of electronic datasets, and returning a result to the user, wherein the result includes one or more datasets of the plurality of electronic datasets with a correlation score greater than or equal to a threshold.

System and method for metadata correlation using natural language processing

An embodiment of the invention provides a method of monitoring for fraudulent activity where a key generating device generates a first key based on a writing profile of a user, and where the key indicates the writing style of the user. The generation of the key includes generating trait scores for writing style traits, where the writing style traits includes an agreeableness trait, a conscientiousness trait, an extraversion trait, an emotional range trait, and an openness trait. The key generating device generates a second key based on a document, where the second key indicates the writing style of the author of the document. A processor compares the first key to the second key to determine the degree of dissimilarity between the writing style of the user and the writing style of the author of the document.

Growing community of inventors

Melbourne, Australia

Lianhua Chi