Language Sample Analysis With TalkBank: An Update and Review

This chapter examines state-of-the-art methods for coding, analyzing, and interpreting discourse-level language data from children and adults with language disorders using the data, tools, and methods provided by the TalkBank system (). These open and free methods have been used for language sample analysis (LSA) with several clinical populations (e.g., child language disorders, stuttering, aphasia, dementia, traumatic brain injury, right hemisphere brain damage), as well as with control participants without communication impairments. We review the six core principles guiding TalkBank, the current shape of the 15 TalkBank databanks, and the different analytic tools provided by TalkBank. We examine automatic TalkBank methods that use ASR (automatic speech recognition), NLP (natural language processing), database technology, statistics in R and Python, and ML (machine learning). The specific tools include corpus analysis methods, LSA profiling systems, online database searches through TalkBank, online browsing through transcripts linked to media, and a new system for online collaborative commentary. These systems provide multimedia access to transcripts from a wide variety of participants with and without language disorders.