Abstract

Speaker: Olesya Razuvayevskaya (NLP group)

Title: Argument mining using argument scheme structures

Abstract: Argument mining from texts in natural language is a challenging task: besides linguistic aspects, domain knowledge is often required together with appropriate forms of inferences to identify arguments. In this talk, I will describe how the structure of argumentation schemes can provide rich information to the task of automatically identifying complex argumentative structures in natural language text. Argumentation schemes are patterns of human reasoning which have been studied extensively in philosophy and psychology. Our results are promising and demonstrate that given the individual proposition types which occur in these schemes, it is possible not only to determine where a scheme is being used, but also the roles played by its component parts.

===============

Speaker: Diana Popescu (NetOS group)

Title: Characterizing Network Latency Impact On Cloud-based Applications Performance

Abstract: Businesses and individuals run increasing numbers of applications using computing resources offered by cloud providers. Well defined performance guarantees are of paramount importance to customers, ensuring lasting collaborations between operators and users. However, the performance of an application depends on the data center conditions and upon the resources committed to an application. Building upon an observation that small network delays may lead to a significant performance degradation, we seek to quantify how latency impacts the application performance for several cloud-based applications. We focus upon the cloud-based applications: DNS server, key-value store and a distributed machine learning systems, to study implementation-based latency-performance. Our results illustrate the impact of latency upon application performance at scales much smaller than typically considered, and we propose a naive model of application performance dependent upon network latency.

===============