On localizing urban events with Instagram

Abstract

Abstract-This paper develops an algorithm that exploits picture-oriented social networks to localize urban events. We choose picture-oriented networks because taking a picture requires physical proximity, thereby revealing the location of the photographed event. Furthermore, most modern cell phones are equipped with GPS, making picture location, and time metadata commonly available. We consider Instagram as the social network of choice and limit ourselves to urban events (noting that the majority of the world population lives in cities). The paper introduces a new adaptive localization algorithm that does not require the user to specify manually tunable parameters. We evaluate the performance of our algorithm for various real-world datasets, comparing it against a few baseline methods. The results show that our method achieves the best recall, the fewest false positives, and the lowest average error in localizing urban events.

Social Fusion: Integrating Twitter and Instagram for Event Monitoring

“…To do so, we first run Instagram-based event detection [11] to get the event clusters. For each cluster, we selected a few random tweets that are relevant and a few few that are not, and generated a frequency distribution of the respective Coherence scores.…”

“…In order to evaluate the performance of our fusion model at event detection we select two individual event detection techniques for each Instagram and Twitter. The Points of Interest (POI) method described in [22] and the Instagram Event localization (InstaLoc) [11] are used for detecting events on Instagram dataset. The Earthquake detection (TweetEvent) [21] and ClariSense [10] are used for detecting events on Twitter dataset.…”

“…[22] is the earliest work that uses Instagram to study the urban social behavior and the city dynamics. In our own recent work [11] we showed how to identify events for urban spaces in an unsupervised way.…”

“…The idea of event detection from Instagram dates back several years [19], [24], [22], [23]. Unlike Twitter, where only only 1.5% of the Twitter data is geo-tagged [18], Instagram has a significantly higher fraction of objects with location information [11]. In our own previous work, we described a system that uses feeds from Twitter [10] (alone) and Instagram [11] 1 (alone) to detect events.…”

“…Unlike Twitter, where only only 1.5% of the Twitter data is geo-tagged [18], Instagram has a significantly higher fraction of objects with location information [11]. In our own previous work, we described a system that uses feeds from Twitter [10] (alone) and Instagram [11] 1 (alone) to detect events. This paper builds on such prior work by offering a novel fusion algorithm that aims to offer a better trade-off between precision and recall of the individual approaches.…”