Prediction of a hotspot pattern in keyword search results. (March 2018)

• Record Type: Journal Article
• Authors: Gao, Jie; Radeva, Axinia; Shen, Chuyao; Wang, Shiqi; Wang, Qianbo; Passonneau, Rebecca J.
• Is Part Of: Computer speech & language. Volume 48(2018)
• Page Start: 80
• Page End: 102
• Publication Date: 2018-03
• Subjects: Self-exciting point process -- Data selection -- Keyword search -- Speech retrieval; Speech processing systems -- Periodicals; Automatic speech recognition -- Periodicals; Computers -- Periodicals; Linguistics -- Periodicals; Speech-Language Pathology -- Periodicals; Traitement automatique de la parole -- Périodiques; Reconnaissance automatique de la parole -- Périodiques; Automatic speech recognition; Speech processing systems; Electronic journals; Periodicals; 006.454
• DOI: 10.1016/j.csl.2017.10.005
• Languages: English
• ISSNs: 0885-2308

Abstract

Highlights: Hotspots of keyword search detections over speech recognition output are identified. Keyword hotspots are modeled as Hawkes process to automatically label data. Prosodic features are used to predict hotspots for word-sized time intervals. Two challenges, class imbalance and disparity of training and test data, are addressed. A novel data selection method with good generalization properties is proposed. Abstract: This paper identifies and models a phenomenon observed across low-resource languages in keyword search results from speech retrieval systems where the speech recognition has high error rate, due to very limited training data. High confidence correct detections (hccd s) of keywords are rare, yet often succeed one another closely in time. We refer to these close sequences ofhccd s as keyword hotspots . The ability to predict keyword hotspots could support speech retrieval, and provide new insights into the behavior of speech recognition systems. We treat hotspot prediction as a binary classification task on all word-sized time intervals in an audio file of a telephone conversation, using prosodic features as predictors. Rare events that follow this pattern are often modeled as a self-exciting point process (sepp ), meaning the occurrence of a rare event excites a following one. To label successive points in time as occurring within a hotspot or not, we fit asepp function to the distribution ofhccd s in the keyword search output. Two major learning challenges are that the size of the positive class is very small, and the training and test data have dissimilar distributions. To address these challenges, we develop a novel data selection framework that chooses training data with good generalization properties. Results exhibit superior generalization performance.