International Business Machines Corporation

International Business Machines Corboration

Mohan Nagraj Dani

Girish Venkatachaliah

Mukesh Kumar Mohania

Anuj Gupta

Sharath C Athrey

Balaji Rishi Sundarajan

L Venkata Subramaniam

Tanveer Afzal Faruquie

Hima Prasad Karanam

Harshavardhan Changappa

Raviprasad Pentakota

Noor Mohammed Ashrafi

Samir Katti

Ramesh G Srinivasan

A computer-implemented system, method and computer program product for capturing data transitions and selecting machine learning models that includes: providing a machine learning model trained with a previous training data set; receiving a new data set; comparing the new data set to the previous data set; and identifying and recording new, removed, and/or changed set of attributes added to the new data set. Future possible data state transitions are generated based upon the present data state; and an implication tree is generated based upon the present data state, the pass-through data states, and the future possible data state transitions. Performance metrics of each node in the implication tree are clustered, the nodes demonstrating a high variance optionally are discarded; reachability scores for remaining nodes are calculated; and a node (representing a machine learning model to run) is selected based upon its reachability score.

System and method to capture data transitions and select optimal data intensive machine learning model

A plurality of liability factors of an asset is identified. Each liability factor impacts a total liability of the asset. A magnitude of a predetermined magnitude scale is identified for each of the plurality of liability factors according to a predetermined magnitude scale. Each magnitude is graphically encoded into a liability image for the asset. A risk score of the asset is determined by analyzing historical correlations between respective magnitudes of the plurality of liability factors using image recognition techniques on the liability image.

Asset assessment via graphical encoding of liability

Embodiments for analyzing language content are provided. An ambiguous sentence is identified. A first word and a second word within the ambiguous sentence are selected. At least one synonym outlier for the first word and at least one synonym outlier for the second word are determined. A phrase is generated utilizing the at least one synonym outlier for the first word and the at least one synonym outlier for the second word. A context for the ambiguous sentence is determined based on the generated phrase.

Methods and systems for improving language processing for ambiguous instances

An approach for training a machine learning model to detect market abuse patterns based on graphical images is disclosed. The approach comprises of creating trade event graphical images based on one or more rules and creating trade risk event graphical images represented by one or more horizontal lines and one or more vertical lines wherein the one or more vertical lines are based on unstructured data. The approach trains a model of a machine learning network to detect market abuse patterns based on the trade event graphical images and the trade risk event graphical images.

Detection of market abuse patterns by artificial intelligence

During application of data quality rules to a data set obtained from a data source, data is retrieved from the data source along with a common set of rules configured to format the retrieved data in a manner in accordance with one or more predefined data quality rules of the common set of rules. At least one predefined data quality rule is adjusted utilizing at least one editable widget to form a modified set of data quality rules adapted for use with a specified application. The modified set of data quality rules is applied to the retrieved data.

Adapting data quality rules based upon user application requirements

Systems, methods, and computer products for optimally managing large rule sets are disclosed. Rule dependencies of rules within a set of rules may be determined as a function of rules execution frequency data generated from applying the rules over a data set. The rules within the set of rules may be clustered into rules clusters based on the determined rule dependencies, in which the rules clusters comprise disjoint subsets of the rules within the set of rules. Cluster frequency data for the rules clusters may be used to arrive at an optimal ordering. Each rule within the set of rules may be assigned a unique identification that may capture an execution order of the rules within the set of rules.

Growing community of inventors

Bengaluru, India

Mohan Nagraj Dani