Humans and machines learning together
Date:
14/10/2020
Organised by:
University of Surrey
Presenter:
Dr. Stuart W. Shulman
Level:
Entry (no or almost no prior knowledge)
Contact:
Venue: Online
Description:
Recent events and emerging trends highlight the need for software that supports rapid and accurate text and social data analytics at scale. One example is the emergence of COVID-19 and related misinformation. Another is the need to identify foreign and domestic efforts to manipulate elections using bots and trolls to amplify propaganda and disinformation. We also see new challenges to academic research using social media data, a rise in suicides and suicidal attempts by military veterans, and an increase in violent attacks on schools and communities that might be preventable through analysis of social media posts. DiscoverText is a decade-old cloud software application that supports teams working to meet these needs by collecting, cleaning and analyzing text and Twitter data. It offers a point-and-click interface to crowdsource the labeling of large data sets. Developed through a decade of U.S. federal funding for basic and applied research, it supports the work of academic research groups, regulatory agencies, non-profit legal teams, and private companies. It is robust in more than 30 languages including the 10 most spoken on earth.
In this webinar, the developer of DiscoverText, Dr Stu Shulman, will discuss the methodological underpinnings of the application, illustrate how build custom machine classifiers for sifting free text, emails, survey responses, Twitter data, RSS feeds, etc., and describe how to reach and substantiate inferences using a theoretical and applied model informed by a decade of interdisciplinary research into the text classification problem.
Cost:
FREE
Website and registration:
https://www.eventbrite.co.uk/e/120027165483
Region:
International
Keywords:
ICT and Software
Related publications and presentations: