Date: 15 February 2021
Time: 16:00 (CET)
Title: Variety of crowdsourcing implementations: ML and business applications
Abstract: The fact that data labeling is one of the crucial bottlenecks in modern AI production is widely known. However, for some reason the potential of Crowdsourcing as an effective method of quality data labeling at a large scale has been underestimated for many years.
On the meeting, Olga Megorskaya, CEO of Toloka, one of the biggest crowdsourcing platforms with more than 9mln registered performers from all over the globe, and the main supplier of labelled data for all ML products at Yandex, will share a 6-years experience in effective use of Crowdsourcing for various real business cases: from Search, Voice assistants and Self-Driving cars to collecting offline data, writing texts and manual testing of software.
Speaker Biography: Initially, Olga was responsible for providing human-labeled data for all AI projects at Yandex. Now Olga is developing Toloka - a global crowdsourcing platform for engineers that allows to integrate data labelling into continous processes of ML production.
She graduated from the Saint Petersburg State University as a specialist in Mathematical Methods and Modeling in Economics. Olga is a co-author of research papers and tutorials on efficient crowdsourcing and quality control at SIGIR, CVPR, KDD, WSDM, and SIGMOD, and led the panel discussion at Crowd Science workshop at NeurIPS'20, a PC member of TheWebConf2021.
(Get Your Toloka Promo Code Here)