System and method for implementing a guided collaboration platform for specific domains
AssigneeKpmg LLP
InventorsJoseph Gwozdz, Timothy J. Cerino, Carly Cullen, David Scripka, Rachel Wagner-Kaiser
Life of this patent
Estimating the term…
Abstract
The invention relates to computer-implemented systems and methods for analyzing and standardizing various types of input data such as structured data, semi-structured data, unstructured data, and images and voice. An embodiment of the present invention relates to a guided collaboration space for domain experts, data scientists and solution managers to build Artificial Intelligence (AI)/Machine Learning (ML) solutions, deploy solutions at scale as well as manage quality and consistency.
Claims20
A computer-implemented system for implementing a collaboration platform that analyzes structured and unstructured data, the system comprising:
- a file ingestion interface that communicates with one or more data sources via a communication network;
- and
- a computer processor coupled to the file ingestion interface and further programmed to perform the steps of:
- ingesting, via the file ingestion interface, unstructured data from the one or more data sources;
- initiating an annotation process that comprises:
- (1) applying a corpus exploration tool to identify a representative set from a corpus that comprises the unstructured data wherein the representative set is relevant to a specific query and (2) applying a rapid annotation tool that applies an artificial intelligence (AI) tool to extract responses relevant to the query as applied to the representative set;
- enabling, via a collaboration interface, a subject matter expert (SME) to perform a verification on the responses extracted from the representative set;
- based on the verification, refining the responses through a feedback process that applies the responses to the annotation process through a machine learning (ML) algorithm;
- and
- responsive to the annotation process, building a model to identify and extract a response relevant to the query as applied to a specified input.
A computer-implemented method for implementing a collaboration platform that analyzes structured and unstructured data, the method comprising the steps of:
- ingesting, via a file ingestion interface, unstructured data from the one or more data sources wherein the file ingestion interface that communicates with one or more data sources via a communication network;
- initiating, via a computer processor, an annotation process that comprises:
- (1) applying a corpus exploration tool to identify a representative set from a corpus that comprises the unstructured data wherein the representative set is relevant to a specific query and (2) applying a rapid annotation tool that applies an artificial intelligence (AI) tool to extract responses relevant to the query as applied to the representative set;
- enabling, via a collaboration interface, a subject matter expert (SME) to perform a verification on the responses extracted from the representative set;
- based on the verification, refining the responses through a feedback process that applies the responses to the annotation process through a machine learning (ML) algorithm;
- and
- responsive to the annotation process, building a model to identify and extract a response relevant to the query as applied to a specified input.
Dependencies are inferred from the claim text.