Real-Time Detection, Tracking, and Monitoring of Automatically Discovered Events in Social Media

Abstract

We introduce ReDites, a system for realtime event detection, tracking, monitoring and visualisation. It is designed to assist Information Analysts in understanding and exploring complex events as they unfold in the world. Events are automatically detected from the Twitter stream. Then those that are categorised as being security-relevant are tracked, geolocated, summarised and visualised for the end-user. Furthermore, the system tracks changes in emotions over events, signalling possible flashpoints or abatement. We demonstrate the capabilities of ReDites using an extended use case from the September 2013 Westgate shooting incident. Through an evaluation of system latencies, we also show that enriched events are made available for users to explore within seconds of that event occurring.

Extracting a Topic Specific Dataset from a Twitter Archive

“…We therefore investigated the best way to extract a topic-specific data set from a previously gathered Twitter archive provided by the ReDites research group [3]. This study investigates the best methods for extracting data relating to the conflict in Syria.…”

Streaming Text Analytics for Real-Time Event Recognition

“…The ReDites real-time event detection and tracking system presented by Osborne et al (2014) focuses on events from the security domain, and uses Twitter as its sole data source. As our system, it is fully automatic, geo-locates events, and supports multiple languages.…”

“…While general text analytics pipelines for event detection focus on detecting and tracking "news topics", such as earthquakes or crisis events (Allan et al, 1998;Osborne et al, 2014), monitoring fine-grained events, such as strikes at a particular facility of a company, or a traffic accident on a specific road crossing, raises additional challenges. It requires more extensive linguistic analysis of texts, including named entity recognition (NER) for non-standard entity types such as roads or facility locations, entity linking (EL) (Dredze et al, 2010) and relation extraction (RE) (Mintz et al, 2009).…”

Online Sentence Novelty Scoring for Topical Document Streams

“…Moreover, every few minutes item churn takes place and the stories of interest are likely to be the ones that appeared in the last couple of hours. As real-time processing on a large scale gains more attention (Osborne et al, 2014), we investigate features that are both effective and efficient, and so could be used in a scalable online novelty scoring engine for making personalized newsfeeds on large web properties like Google News and Yahoo News.…”