Abstract: 

Natural language processing has achieved remarkable progress over the last decade, and many companies now rely on NLP-based insights to drive their business intelligence. One of the more challenging goals is to elevate the capabilities of AI models from identifying companies and events in text-based news to building clusters that define a topic, to discerning trends around those topics from which financial professionals can derive actionable insights. Up until recently the last step of that process relied on manual intervention, but recent developments involving deep learning models now make the seamless automation of topic modeling and trend detection within reach. Completing that last step would allow for constructing a context of entities and topics around events using a knowledge graph, and generating an awareness of trends affecting a universe of interests for human users, which in turns would enable us to build narratives that can be plotted on a timeline and describe the evolution of a particular relationship. In this session, we’ll unpack the contributions of these new models, and explore the consequences of the improvements to industries like supply chain risk management, investment management and business intelligence.

Bio: 

Marko is the Head of Quantitative Research for the Americas at RavenPack with over 10 years of experience in the finance industry. He focuses on exploring novel approaches and techniques for combining fundamental drivers with big data quantitative frameworks to identify alpha opportunities from a wide universe of securities across multiple asset classes. Previously, as the head trader/investment analyst at an event-driven hedge fund in New York, he was responsible for macro research, idea generation and risk management. Marko has experience in utilizing quantitative methods in portfolio construction, developing hedging strategies and trading structured derivative instruments.