Sasa Petrovic

Scalable distributed event detection for Twitter

Social media streams, such as Twitter, have shown themselves to be useful sources of real-time information about what is happening in the world. Automatic detection and tracking of events identified in these streams have a variety of real-world applications, e.g. identifying and automatically reporting road accidents for emergency services. However, to be useful, events need to be identified within the stream with a very low latency. This is challenging due to the high volume of posts within these social streams. In this paper, we propose a novel event detection approach that can both effectively detect events within social streams like Twitter and can scale to thousands of posts every second. Through experimentation on a large Twitter dataset, we show that our approach can process the equivalent to the full Twitter Firehose stream, while maintaining event detection accuracy and outperforming an alternative distributed event detection system.

Extending lexical association measures for collocation extraction

Collocations are linguistic phenomena that occur when two or more words appear together more often than by chance and whose meaning often cannot be inferred from the meanings of its parts. As collocations have found many applications in the fields of natural language processing, information retrieval, and text mining, extracting them from large corpora has been the focus of many studies over the past few years. In this paper, we introduce the notion of an extension pattern, a formalization of the idea of extending lexical association measures (AMs) defined for bigrams. An extension pattern provides a measure-independent way of extending AMs for extracting collocations of arbitrary length. We define different extension patterns and compare them on a task of extracting collocations from a newspaper corpus. We show that the stopword-sensitive extension patterns we propose outperform other extensions, which indicates that AMs could benefit by taking into account linguistic information about an n-grams part-of-speech pattern.

Textual features for corpus visualization using correspondence analysis

Explorative data analysis in text mining essentially relies on effective visualization techniques which can expose hidden relationships among documents and reveal correspondence between documents and their features. In text mining, the documents are most often represented by feature vectors of very high dimensions, requiring dimensionality reduction to obtain visual projections in two- or three-dimensional space. Correspondence analysis is an unsupervised approach that allows for construction of low-dimensional projection space with simultaneous placement of both documents and features, making it ideal for explorative analysis in text mining. Its present use, however, has been limited to word-based features. In this paper, we investigate how this particular document representation compares to the representation with letter n-grams and word n-grams, and find that these alternative representations yield better results in separating documents of different class. We perform our experimental analysis on a bilingual Croatian-English parallel corpus, allowing us to additionally explore the impact of features in different languages on the quality of visualizations.

Evolving New Lexical Association Measures Using Genetic Programming

Automatic extraction of collocations from large corpora has been the focus of many research efforts. Most approaches concentrate on improving and combining known lexical association measures. In this paper, we describe a genetic programming approach for evolving new association measures, which is not limited to any specific language, corpus, or type of collocation. Our preliminary experimental results show that the evolved measures outperform three known association measures.