Language and Computation Seminar Series

• Steven Zimmerman, University of Essex, UK
• 05/19 - 3pm, TBA
• Abstract

The main focus of my PhD considers methods to assist information retrieval (IR) system users into making better decisions. Nudging and Boosting are the methods approaches I am most interested and will briefly introduce both. My research is closely related to the nutrition label work presented to our group in January. During this time, I will also overview methods and findings from my recently completed study that will be published in the upcoming CHIIR conference. I will include methods and results from a recent experiment that will be This study looks at Nudging users towards privacy protection in the health search domain, with the aim of understanding impacts on their decisions.

Time permitting, I will give an overview of my early PhD research into hate speech detection (and the difficulties surrounding this task), which was presented at LREC 2018.

Short Bio

Steven Zimmerman is a PhD candidate at the University of Essex, under the advisement of Udo Kruschwitz and Chris Fox. Steven, Chris and Udo, are part of the Human Rights Big Data and Technology project, also based at the University.

• James Brill
• TBA - 3pm, TBA
• Abstract

Signal is an AI-powered platform for media intelligence. Media analysts use Signal to analyse news and web content and carry out a wide variety of tasks. These include reputation management, business lead generation, and conducting due diligence. All these tasks require the media analysts to digest large number of documents about a specific person or organisation, which is time-consuming and repetitive due to the multitude of irrelevant information. At Signal, we recently undertook a research project whose goal was to save media analysts time by summarising the most relevant information about a person or organisation of interest. This talk will provide a walkthrough of the research project, including the solution, company impact, and learnings.

Short Bio

James is a product data scientist and KTP associate at Signal. He recently graduated with a masters in machine learning and data science from the University of Bristol with his thesis exploring the generation and detection of fake news.