Extracting hot spots of topics from time-stamped documents

Abstract

Identifying time periods with a burst of activities related to a topic has been an important problem in analyzing time-stamped documents. In this paper, we propose an approach to extract a hot spot of a given topic in a time-stamped document set. Topics can be basic, containing a simple list of keywords, or complex. Logical relationships such as and, or, and not are used to build complex topics from basic topics. A concept of presence measure of a topic based on fuzzy set theory is introduced to compute the amount of information related to the topic in the document set. Each interval in the time period of the document set is associated with a numeric value which we call the discrepancy score. A high discrepancy score indicates that the documents in the time interval are more focused on the topic than those outside of the time interval. A hot spot of a given topic is defined as a time interval with the highest discrepancy score. We first describe a naive implementation for extracting hot spots. We then construct an algorithm called EHE (Efficient Hot Spot Extraction) using several efficient strategies to improve performance. We also introduce the notion of a topic DAG to facilitate an efficient computation of presence measures of complex topics. The proposed approach is illustrated by several experiments on a subset of the TDT-Pilot Corpus and DBLP conference data set. The experiments show that the proposed EHE algorithm significantly outperforms the naive one, and the extracted hot spots of given topics are meaningful.

Creation of knowledge-added concept maps: time augmention via pairwise temporal analysis

“…Finally, a third limitation of most ETD systems relates to the use of a unitary static monolithic text corpus from human-maintained indexed databases, such as INSPEC, topic detection and tracking (TDT) or COMPENDEX (Nowell et al, 1997;Lent et al, 1997;Swan and Jensen, 2000;Wong et al, 2000;Kumaran and Allan, 2004;Mei and Zhai, 2005;Zhang et al, 2007;Subasic and Berendt, 2010;Chen and Chundi, 2011). A closed static textual data corpus suffers from limited diversity, variety and richness and must be periodically refreshed, imposing major drawbacks, such as data coverage and indexer effect (Alexa, 1997;Zweigenbaum et al, 2001;Banko and Brill, 2001;Keller and Lapata, 2003).…”

Explore the research front of a specific research theme based on a novel technique of enhanced co-word analysis

“…Sun et al [16] proposed a novel algorithm for BTD within multiple data streams via a network-based method. Chen and Chundi [17] found a technique to extract topic hot spots from text based on a time-stamped sequence. Takahashi et al [4] proposed a method to discover emerging topics from social streams based on an analysis of Link-Anomaly Detection.…”

Constructing Temporal Equivalence Partitionings for Keyword Sets

“…Although identifying temporal information from text documents is an active area of research (Chundi and Rosenkrantz, 2006;Chen and Chundi, 2011), to the best of our knowledge, identifying temporally changing keyword relationships from document sets is still an emerging field. In and , Kage and Sumiya define several temporal relationships among keywords, such as co-occurring, ordered, and define an approach to determine these relationships.…”