Sentence segmentation of aphasic speech

Human Language Technologies: The 2015 Annual Conference of the North American Chapter of the ACL, pages 862–871, Denver, Colorado, May 31 – June 5, 2015. c⃝2015 Association for Computational Linguistics Sentence segmentation of aphasic speech Kathleen C. Fraser1,3, Naama Ben-David1, Graeme Hirst1, Naida L. Graham2,3, Elizabeth Rochon2,3 1Department of Computer Science, University of Toronto, Toronto, Canada 2Department of Speech-Language Pathology, University of Toronto, Toronto, Canada 3Toronto Rehabilitation Institute, Toronto, Canada {kfraser,naama,gh}@cs.toronto.edu, {naida.graham,elizabeth.rochon}@utoronto.ca Abstract Automatic analysis of impaired speech for screening or diagnosis is a growing research field; however there are still many barriers to a fully automated approach. When automatic speech recognition is used to obtain the speech transcripts, sentence boundaries must be in- serted before most measures of syntactic com- plexity can be computed. In this paper, we consider how language impairments can affect segmentation methods, and compare the re- sults of computing syntactic complexity met- rics on automatically and manually segmented transcripts. We find that the important bound- ary indicators and the resulting segmentation accuracy can vary depending on the type of impairment observed, but that results on pa- tient data are generally similar to control data. We also find that a number of syntactic com- plexity metrics are robust to the types of seg- mentation errors that are typically made. 1 Introduction The automatic analysis of speech samples is a promising direction for the screening and diagno- sis of cognitive impairments. For example, recent studies have shown that machine learning classifiers trained on speech and language features can detect, with reasonably high accuracy, whether a speaker has mild cognitive impairment (Roark et al., 2011), frontotemporal lobar degeneration (Pakhomov et al., 2010b), primary progressive aphasia (Fraser et al., 2014), or Alzheimer’s disease (Orimaye et al., 2014; Thomas et al., 2005). These studies used manually transcribed samples of patient speech; however, it is turning to politics for al gore and george w bush another day of rehearsal in just over forty eight hours the two men will face off in their first of three debates for the first time voters will get a live unfiltered view of them together Turning to politics, for Al Gore and George W Bush an- other day of rehearsal. In just over forty-eight hours the two men will face off in their first of three debates. For the first time, voters will get a live, unfiltered view of them to- gether. Figure 1: ASR text before and after processing. clear that for such systems to be practical in the real world they must use automatic speech recognition (ASR). One issue that arises with ASR is the intro- duction of word recognition errors: insertions, dele- tions, and substitutions. This problem as it relates to impaired speech has been considered elsewhere (Jar- rold et al., 2014; Fraser et al., 2013; Rudzicz et al., 2014), although more work is needed. Another is- sue, which we address here, is how ASR transcripts are divided into sentences. The raw output from an ASR system is generally a stream of …

Content truncated to 500 words