Robust Classification of Dialog Acts from the Transcription of Utterances

Abstract

This paper presents a robust classification of dialog acts from text utterances. Two different types, namely, bag-of-words and syntactic relationship among words, were used to extract the discourse level features from the transcript of utterances. Subsequently a number of feature mining methods have been used to identify the most relevant features and their roles in classifying dialog acts. The selected features are used to learn the underlying models of dialog acts using a number of existing machine learning algorithms from the WEKA toolbox. Empirical analyses using the HCRC Map Task Corpus dialog data was conducted to evaluate the performance of the proposed approach.

Discriminative histogram taxonomy features for snake species identification

“…In order to find the top features from the complete database following 12 Attribute Elevators are used: ChiSquared AttributeEval [12], CfsSubsetEval [13], ConsistencySubsetEval [14], FilteredAttributeEval [15], FilteredSubsetEval [16], GainRatioAttributeEval [17], InfoGainAttributeEval [18], OneRAttributeEval [19], PrincipalComponents [20], ReliefFAttributeEval [21], SVMAttributeEval [19], SymmetricalUncertAttributeEva [19], along with combination of certain search methods [21,22] like Genetic Search, Greedy Stepwise, Linear Forward Selection, Rank Search, Scatter Search, Subset Size Forward Selection and Ranker. The histogram of the feature counts from these attribute elevators is then plotted to get the ranking of the taxonomically relevant features that are most useful for the classification as shown in Figure 2.…”