| CPC G06Q 10/0635 (2013.01) [G06F 16/215 (2019.01); G06F 40/242 (2020.01); G06N 5/04 (2013.01)] | 17 Claims |
|
1. A method for identifying and mitigating risks in a network, the method comprising:
receiving, by one or more processors, a historic dataset and a current dataset from one or more data sources via the network, wherein each of the historic dataset and the current dataset comprise data associated with one or more entities, and wherein each of the historic dataset and the current dataset comprise a plurality of data items obtained from the one or more data sources;
generating, by the one or more processors, a pruned dataset by using a first machine learning model for pruning the current dataset based on a sentiment of each data item of the plurality of data items of the current dataset, wherein the sentiment for each data item of the plurality of data items of the current dataset is determined based on executing a set of rules against the current dataset, wherein
the set of rules includes a dictionary customized to the one or more entities, and
the sentiment for a data item of the plurality of data items of the current data set is determined by:
associating a corresponding value to each term of a plurality of terms of the data item based on analysis of the data item with the dictionary; and
determining the sentiment of the data item based on sum of corresponding values of the plurality of terms of the data item;
dividing, by a modelling engine, the historic dataset into a training dataset and a testing dataset;
evaluating, by the modelling engine, performance of a second machine learning model based on the testing dataset;
comparing, by the modelling engine, the performance of the second machine learning model with a threshold performance level to determine a result;
iteratively updating, by the modelling engine, the second machine learning model based on the result indicating that the second machine learning model is below the threshold performance level, wherein the iteration update of the second machine learning model comprises:
adjusting the training dataset and the testing dataset in a plurality of cycles; and
training and evaluating the second machine learning model based on the adjusted training dataset and the adjusted testing dataset in each of the plurality of cycles,
wherein the second machine learning model is iteratively updated to generate an updated second machine learning model, and
wherein the updated second machine learning model satisfies the threshold performance level;
evaluating, by the modelling engine, the pruned dataset against the updated second machine learning model to produce a set of risk categorizations, wherein the set of risk categorizations associate a risk category with each data item of the plurality of data items;
generating, by the one or more processors, scoring metrics for each of the plurality of data items based at least in part on the evaluating,
wherein the scoring metrics account for an impact of dependencies and capabilities of the one or more entities on the risks corresponding to the set of risk categorizations,
wherein the scoring metrics comprise risk scores, capability scores, and dependency scores,
wherein the risk scores indicate the risks related to the one or more entities,
wherein the capability scores indicate capabilities of the one or more entities, and
wherein the dependency scores evaluate aspects of a relationship between at least one of:
two or more different entities of the one or more entities, or
different portions of a single entity of the one or more entities;
creating, by the one or more processors, a data structure based on the scoring metrics,
wherein the data structure is configured to quantify the risks identified for each entity of the one or more entities and to identify actions configured to mitigate the risks for each entity,
wherein the data structure quantifies:
the risks corresponding to the risk scores along a first axis of a three-dimensional space,
the dependencies corresponding to the dependency scores on a second axis of the three-dimensional space, and
the capabilities corresponding to the capability scores along a third axis of the three-dimensional space; and
displaying, at a display device, the data structure to a user via a graphical user interface, wherein the graphical user interface comprises interactive elements enabling the user to initiate one or more actions identified in the data structure.
|