Who, What, Where, When and How: Analyzing Documents with NLP, ML and Graphs

Please join us in March as we return to FINRA to learn how they are using machine learning to analyze hundreds of thousands of financial documents.

Agenda

6:30 PM -- Networking & Food

7:00 PM -- Greetings

7:05 PM -- Using NLP, ML & Graph Databases to Automate a Documents Review Process - Dmytro Dolgopolov

Location

FINRA
9513 Key W Ave, Rockville, MD 20850

Please bring an ID that matches your registered name.

Parking

There are parking lots in front and to the side of the building.

Talks

Using NLP, ML & Graph Databases to Automate a Documents Review Process
FINRA’s investigators and analysts review hundreds of thousands of various documents each year received from stock brokers and investors. Risk-based approach to determine whether and when to review the documents is in itself a risk. Finding information about the Who, What, Where, When and How contained in hundreds pages long documents is labor intensive.
FINRA employs various advancements in AI to automate and assist with review of the documents that have no structure. Using Natural Language Processing, Machine Learning and Graph Databases allows extracting crucial information as well as making key business decision with higher level of efficiency and confidence.

Speakers

Dmytro Dolgopolov, Senior Director, Content Services and Analytics, Dmytro is a Senior Technology Leader with over 20 years of experience in delivering software solutions. The last several years have been focusing on storing, discovering and analyzing various regulatory text corpora. Currently leading the team of researches and engineers that apply various text analytics techniques for regulatory review of unstructured content. Presented at various international conferences including AWS re:Invent.

Company

FINRA is dedicated to investor protection and market integrity through effective and efficient regulation of broker-dealers. More information can be found at finra.org