On fine-grained geolocalisation of tweets and real-time traffic incident detection

Abstract

Recently, geolocalisation of tweets has become important for a wide range of real-time applications, including real-time event detection, topic detection or disaster and emergency analysis. However, the number of relevant geotagged tweets available to enable such tasks remains insufficient. To overcome this limitation, predicting the location of non-geotagged tweets, while challenging, can increase the sample of geotagged data and has consequences for a wide range of applications. In this paper, we propose a location inference method that utilises a ranking approach combined with a majority voting of tweets, where each vote is weighted based on evidence gathered from the ranking. Using geotagged tweets from two cities, Chicago and New York (USA), our experimental results demonstrate that our method (statistically) significantly outperforms state-ofthe-art baselines in terms of accuracy and error distance, in both cities, with the cost of decreased coverage. Finally, we investigated the applicability of our method in a real-time scenario by means of a traffic incident detection task. Our analysis shows that our fine-grained geolocalisation method can overcome the limitations of geotagged tweets and precisely map incident-related tweets at the real location of the incident.

Utilizing microblogs for assisting post-disaster relief operations via matching resource needs and availabilities

“…There has been a lot of recent work in utilizing Online Social Media (OSM) to facilitate post-disaster relief operations -see [1,8,9] for some recent surveys on this topic. For instance, there have been works on classifying situational and non-situational information [3,10], location inferencing from social media posts during disasters [11,12,13], early detection of rumours from social media posts [14], emergency information diffusion on social media during crises [15], event detection [16], extraction of event-specific informative tweets during disaster [17] and so on. Recent works have exemplified social media's ability to disseminate disaster information between institutional and non-institutional volunteers [18] and their use to reinforce the role of different stakeholders [19].…”

“…Additionally, we utilize a list of common suffixes of location names to recognize locations. The suffix list -a part of which is shown in Table 4 -comprises different naming conventions for landforms 13 , roads 14 15 , buildings 16 and towns.…”

Comparing Methods to Retrieve Tweets: a Sentiment Approach

“…These new sources of data, together with the Internet, are also used, from a practical perspective, in order to support decision processes in several fields. In some of these cases researchers require information that is fully geolocalized: this happens, for example, monitoring socio-demographic phenomena (Jashinsky et al, 2014), in disaster management (de Bruijn et al, 2017) or in transportation planning studies (Paule et al, 2019). In this framework, one of the biggest problem is that tweets with a geographical information are just a small fraction of the total (Middleton et al, 2018).…”

Processing social media in real-time

“…This special issue covers a wide range of applications that rely on real-time processing of social media, including event detection [8,14,7], cybersecurity [10], opinion mining [5] and automatic geo-localization [7]. The diversity of submissions received shows the need for furthering research in processing social media streams in a (near) real-time.…”