Ai training data development Initiatives announced by cogito

Expert System announced today the latest enhancements to Cogito®, its Artificial Intelligence platform that understands textual information and automatically processes natural language.

Shipping at the end of April, the new version of Cogito (14.4) delivers key updates in the areas of knowledge graphs, machine learning and RPA to help organizations accelerate and simplify the adoption of AI in their business workflows.

Cogito® Knowledge Graph

Cogito® 14.4 enables users to more easily customize its Knowledge Graph of approximately 350,000 concepts connected by 2.8 Million relationships and lets them import targeted knowledge from any sources (such as company repositories Wikipedia or Geonames) in only a few clicks, enabling the platform to resolve references to real-world entities (such as people, companies, locations) and to link them to knowledge repositories by using standardized identifiers.

Active Machine Learning Workflow

Cogito 14.4 also extends its Natural Language Processing (NLP) extraction pipeline with a new active learning workflow that accelerates machine-learning-based analytics projects. Because machine learning relies on manually annotated datasets, the early phases of such projects are extremely time-consuming. Through an intuitive web application, Cogito 14.4’s active learning workflow enables end-users to visualize the quality of extraction and provide feedback to the engine, which iteratively retrains the engine to reach the user’s quality goals, thus reducing the amount of manual annotation needed. With all steps of this process accessible from a single, convenient, interface, Cogito 14.4 also enables a larger participation of subject matter experts with limited machine learning expertise in the development phase of machine learning models.

Robotic Process Automation (RPA) Connector

Cogito 14.4 includes a Robotic Process Automation (RPA) connector that extends the use of RPA bots into process automation leveraging knowledge (and not only structured data) as well as requiring human-like judgement. The Cogito RPA Connector leverages deep contextual understanding to extract precise data from unstructured business documents. This enables bots to intelligently classify and route mainstream cases for straight-through processing and to flag suspicious patterns, exceptions and high-risk cases for enhanced review by human operators. This paves the way towards genuinely intelligent process automation, accelerating business workflows and enabling organizations to focus their teams on the steps of their processes with the highest added-value.

“Text analysis, natural language understanding and machine learning are sometimes presented as alternatives but in reality they complement each other. Semantics is ideal to identify meaning in context while machine learning is a useful technique that can save good amounts of time in some scenarios,” said Marco Varone, Expert System‘s President and CTO. “With this new release, we confirm our commitment to supporting our partners’ and clients’ requirements, while encouraging the adoption of Artificial Intelligence across all industries.”

To continue effectively addressing customers’ growing security requirements, the new release of Cogito also embeds hardened data exchange protocols into its external and internal APIs. Cogito can therefore be confidently applied to applications requiring high data security, such as in Insurance, Banking, Government and Healthcare.

BlogofInnovation.com