Search

At present, most of the topic detection approaches are not accurate and efficient enough. In this paper, we proposed a new topic detection method (TPIC) based on an incremental clustering algorithm. It employs a self-refinement process of discriminative feature identification and a term reweighting algorithm to accurately cluster the given documents which discuss the same topic. To be efficient, the ˵aging″ nature of topics is used to precluster stories. To automatically detect the true number of topics, Bayesian Information Criterion (BIC) is used to estimate the true number of topics. Experimental results on Linguistic Data Consortium (LDC) datasets TDT4 show that the proposed method can improve both the efficiency and accuracy, compared to other methods. [ABSTRACT FROM AUTHOR]

New event detection (NED) involves monitoring one or multiple news streams to detect the stories that report on new events. With the overwhelming volume of news available today, NED has become a challenging task. In this paper, we proposed a new NED model based on incremental PLSA(IPLSA), and it can handle new document arriving in a stream and update parameters with less time complexity. Moreover, to avoid the limitation of TF-IDF method, a new approach of term reweighting is proposed. By dynamically exploiting importance of documents in discrimination of terms and documents΄ topic information, this approach is more accurate. Experimental results on Linguistic Data Consortium (LDC) datasets TDT4 show that the proposed model can improve both recall and precision of NED task significantly, compared to the baseline system and other existing systems. [ABSTRACT FROM AUTHOR]

This paper presents the design of a new interface for interactive Topic Detection and Tracking (TDT) called Ievent. It is composed of 3 main views; a Cluster View, a Document View, and a Named Entity View, supporting the user in identifying new events and tracking them in a news stream. The interface has also been designed to test the usefulness in interactive TDT of named entity recognition. We report some initial findings from a user study on the effectiveness of our novel interface. [ABSTRACT FROM AUTHOR]