FloDusTA: Saudi Tweets Dataset for Flood, Dust Storm, and Traffic Accident Events

Proceedings of the 12th Conference on Language Resources and Evaluation (LREC 2020), pages 1391–1396 Marseille, 11–16 May 2020 c⃝ European Language Resources Association (ELRA), licensed under CC-BY-NC 1391 FloDusTA: Saudi Tweets Dataset for Flood, Dust Storm, and Traffic Accident Events Btool Hamoui, Mourad Mars, Khaled Almotairi College of Computer and Information System, Common First Year Deanship Umm AL-Qura University, Saudi Arabia s43680523@st.uqu.edu.sa, {msmars, khmotairi}@ uqu.edu.sa Abstract The rise of social media platforms makes it a valuable information source of recent events and users’ perspective towards them. Twitter has been one of the most important communication platforms in recent years. Event detection, one of the information extraction aspects, involves identifying specified types of events in the text. Detecting events from tweets can help to predict real-world events precisely. A serious challenge that faces Arabic event detection is the lack of Arabic datasets that can be exploited in detecting events. This paper will describe FloDusTA, which is a dataset of tweets that we have built for the purpose of developing an event detection system. The dataset contains tweets written in both Modern Standard Arabic and Saudi dialect. The process of building the dataset starting from tweets collection to annotation by human annotators will be present. The tweets are labeled with four labels: flood, dust storm, traffic accident, and non-event. The dataset was tested for classification and the result was strongly encouraging. Keywords: Arabic tweets, Saudi dialect, Twitter, Event detection, Classification. 1. Introduction In recent years, social media platforms in the Arabic region have been evolving rapidly. Twitter provides an easy form of communication that enables users to share information about their activities, opinions, feelings, and views about a wide variety of social events. It has been a great platform to disseminate events as they happen and released immediately, even before they are announced in traditional media. Tweets’ contents have become major sources for extracting information about real-world events. Critical events such as violence, disasters, fires, and traffic accidents that need emergency awareness require an extreme effort to detect and track. Twitter users’ posts have been utilized as data provider to detect high-risk events with their locations, such as earthquakes (Sakaki et al., 2010), Traffic incidents (Gu et al., 2016) and floods (Arthur et al., 2018). An earlier work done by Sakaki et al. (2010) predicted and detected the location of an earthquake in Japan more quickly than the Japan Meteorological Agency. (Gu et al., 2016) identified five categories of traffic incidents in the city of Pittsburgh and Philadelphia (USA) using twitter data. A recent study by Arthur et al. (2018) utilized tweets to locate and detect flood in the UK. Recently, event detection has been considered an active area of researches due to the widespread availability of data in social media. However, researches about event detection on Twitter applying it on Arabic is hampered by the lack of datasets that could be used to design and develop an event detection system. Until now, the dataset of (Almerekhi et al., 2016) and (Alhelbawy …

Content truncated to 500 words