Bursty event detection from collaborative tags

Abstract

Collaborative tagging have emerged as a ubiquitous way to annotate and organize online resources. As a kind of descriptive keyword, large amount of tags are created and associated to multiple types of resources, e.g., web pages, photos, videos and tweets. Users' tagging actions over time reflect their changing interests. Monitoring and analyzing the temporal patterns of tags can provide important insights to trace hot topics on the web. Existing work focuses on deriving temporal patterns for individual tags. However, there exist remarkable correlations among tags assigned to online resources. In this paper, we propose a new approach to detect bursty tagging event, which captures the relations among a group of correlated tags where the tags are either bursty or associated with bursty tag co-occurrence. This kind of bursty tagging event generally corresponds to a real life event. It profiles the events with more representative and comprehensible clues. The proposed approach is divided into three stages. We exploit the sliding time intervals to extract bursty features as the first step, and then adopt graph clustering techniques to group bursty features into meaningful bursty events. We discuss the choice of similarity and granularity for event detection. After that, we further utilize an automatically generated tag taxonomy to organize bursty events to facilitate the burst oriented 172 World Wide Web (2012) 15:171-195 navigation and analysis. The experimental study on a large real data set demonstrates the superiority of our new approach.

Clustering Based Topic Events Detection on Text Stream

“…The event detection generally involves with several research domains, such as topic detection and tracking [7,8,9], text clustering [10,11,12,13,14] and temporal data analysis. The topic detection and tracking based approach is originally proposed to discover the topic hidden in the stream of news stories [7].…”

Event Detection in Multiple Webpages based on Comprehensive Dimension Matching and Co-occurrence Constraint

“…Information retrieval identifies similar changes in the relationship [4] between certain entities, such as simultaneously rise and decline, to detect events. Social tagging [5,6] discovers burst tags in the same period to detect a few special events. Manoj [7] found real-time events according to high frequency words in micro-blogs reported by the same author.…”

Estimating the semantic type of events using location features from Flickr

“…For this purpose, a Naive Bayes classifier was trained on photos associated with events of known types. Yao et al [16] detected events using the tagging history of the social bookmarking webservice Del.icio.us. The authors organized the detected events by mapping them to a hierarchy of semantic types, i.e.…”

“…The textual content associated with a photo consists of a set of tags, a title and a description. In previous work, the textual content of social media documents has already been used to classify events [6,16]. In this 'bag-of-words' approach, a vector describing an event e ∈ E is constructed, whose components are associated with a word that appears in dictionary W .…”