TA-COS 2018 : 2nd Workshop on Text Analytics for Cybersecurity and Online Safety : Proceedings
Els Lefever, Bart Desmet, Guy De Pauw · Ghent University Academic Bibliography (Ghent University) · 2018
Text analytics technologies are being widely used as components in Big Data applications, allowing for the extraction of different types of information from large volumes of text, including purely factual information ("traditional" text mining), subjective information (sentiment mining) and even metadata (e.g.author profiling).A growing number of research efforts is now investigating the applicability of these techniques for cybersecurity purposes.Many applications are using text analytics techniques to provide a safer and more pleasant online experience, by detecting unwanted content and behavior on the Internet.Other text analytics approaches attempt to detect illegal activity on online networks or monitor social media against the background of real-life threats.The second run of the workshop on Text Analytics for Cybersecurity and Online Safety (TA-COS 2018) aims to bring together researchers that have an active interest in the development and application of text analytics systems in the broad context of cybersecurity.We were interested in research papers on text analytics and text mining approaches that (1) help reduce the exposure to harmful content on the Internet, (2) detect illegal online activities and (3) monitor user-generated content in the context of real-life security threats.Following our call for papers, we received papers on a wide range of topics and with the help of our varied team of reviewers were able to select the most relevant and most interesting contributions.We are very pleased to present a wide variety of topics of the accepted papers for the workshop.Two papers deal with identifying hate speech on social media: Sirihattasak et al. present an annotated corpus and classification experiments for toxic messages in Thai tweets, while Isbister et al. present a case study on monitoring targeted hate in Swedisch online text.Alshehri et al. present a dataset of adult content in Arabic Twitter and provide an in-depth analyses of this data.The fourth paper deals with targeted email attacks.Das and Verma propose a system for advanced email masquerading attacks using Natural Language Generation techniques.We are furthermore very pleased to be able to kick off our workshop with a keynote lecture by Pierre Lison, who is a Senior Research Scientist at Norsk Regnesentral (Norwegian Computing Center), a contract-funded research institute located in Oslo, Norway.He will present research on data-driven models of reputation in cyber-security.We are sure that the presentations at TA-COS 2018 will trigger fruitful discussions and will help foster the awareness of the increasingly important role text analytics can play in cybersecurity applications.