On the collective classification of email "speech acts"

Abstract

We consider classification of email messages as to whether or not they contain certain "email acts", such as a request or a commitment. We show that exploiting the sequential correlation among email messages in the same thread can improve email-act classification. More specifically, we describe a new textclassification algorithm based on a dependency-network based collective classification method, in which the local classifiers are maximum entropy models based on words and certain relational features. We show that statistically significant improvements over a bag-of-words baseline classifier can be obtained for some, but not all, email-act classes. Performance improvement obtained by collective classification is appears to be consistent across email acts suggested by prior speech-act theory.

Collective Classification in Network Data

“…Yang, Slattery, & Ghani [63] conducted an in-depth investigation over multiple datasets commonly used for document classification experiments and identified different patterns. Since then, collective classification has also been applied to various other applications such as part-of-speech tagging [33], classification of hypertext documents using hyperlinks [55], link prediction in friend-of-a-friend networks [56], optical character recognition [58], entity resolution in sensor networks [9], predicting disulphide bonds in protein molecules [57], segmentation of 3D scan data [2] and classification of email "speech acts" [7].…”

“…For instance, in the webpage classification problem where webpages are interconnected with hyperlinks and the task is to assign each webpage with a label that best indicates its topic, it is common to assume that the labels on interconnected webpages are correlated. Such interconnections occur naturally in data from a variety of applications such as bibliographic data [10,16], email networks [7] and social networks [41]. Traditional classification techniques would ignore the correlations represented by these interconnections and would be hard pressed to produce the classification accuracies possible using a collective classification approach.…”

Semi-Automated Dialogue Act Classification for Situated Social Agents in Games

“…There have been numerous approaches to automatically classifying speech acts, including neural network classification [12], maximum entropy model classification [1], and Hidden Markov Model (HMM) speech act classification [21]. Our classifier is composed of three independent HMM classifiers, one for each axis (speech act, content, and referent).…”

Semanta – Semantic Email in Action

“…We also apply Speech Act Theory [6] to provide more adequate representation of email action items. Although the same theory was applied to email beforehand -particularly to address the problems of email task management [7][8] and email classification [9][10] [11]; it was never combined with semantic technologies. We represent action items by speech act instances from within the sMail Ontology 2 .…”