Fellow: ESR8
Host institution: Radboud Universiteit Nijmegen (RUN), Netherlands
Primary supervisor: Assoc. Prof. Helmer Strik
Title (WP): Developing valid measures of pathological speech intelligibility: human ratings and automatic scores
Objectives: Speech disorders that lead to decreased speech intelligibility have severe consequences for the patients and can affect their Quality of Life as they run the risk of losing contact with friends and relatives and eventually becoming isolated from society. Intelligibility can be improved by means of speech therapy. However, little is known about which deviations in pathological speech most affect intelligibility, and how intelligibility can best be improved. Furthermore, measuring intelligibility is complex, since there are no standard procedures, and it is time consuming as it requires a lot of manual work. This project will study the following novel aspects: which deviations in pathological speech have most impact on intelligibility, how therapy can best improve intelligibility, what are good procedures for measuring intelligibility, and how can the work load in measuring intelligibility be reduced by making use of software tools.
.
Additional notes:
The goal of speech training, in the context of therapy [e-Health] or language learning [e-Learning, CALL], nowadays usually is to increase intelligibility. After all, by increasing their intelligibility people will be better able to communicate with others, which generally is their main goal. Therefore, tests are often carried out to determine how intelligible somebody is; and in order to assess how effective an approach, a treatment, is, it is useful to measure the intelligibility before and after the treatment and compare them.
. But intelligibility is a complex construct, and measuring intelligibility is not straightforward. At the moment there is no standard procedure to measure intelligibility, in the fields of speech therapy and language learning different procedures are used to measure it. And current procedures generally require substantial amounts of manual work, making them time-consuming. In project ESR8 of TAPAS, we want to compare these different procedures, and also look at (computational) models of intelligibility, in order to study what the advantages and disadvantages of the different approaches are. In addition, we will study the different types of speech material that can be used in intelligibility tests, e.g., isolated words (either nonsense words or real words), or utterances (either semantically unpredictable sentences or real utterances).
. An important aspect is to study to what extent intelligibility can be measured (semi-)automatically. If it becomes possible to measure intelligibility automatically, to a large extent, it will become easier and less time-consuming to often measure intelligibility, before and after a treatment (pre- and post-test), but also regularly during a treatment. If such [semi-]automatic procedures for measuring intelligibility exist, this also creates new possibilities to study what the most important factors are affecting intelligibility, and what kinds of training work best for increasing intelligibility.
. Our group has been and is involved in many research projects, esp. on e-Health and e-Learning. In these projects we have developed and tested e-Health and e-Learning applications (e.g. serious games), which were and still can be used to collect additional data, and to carry out experiments on speech training. We already have speech data (pathological and non-native speech) with annotations, e.g. subjective intelligibility scores.
.    Linguistics in Nijmegen has recently been ranked nr. 1 in the world. We are a strong language and speech technology group at the department of Linguistics. We thus have the advantage of being able to combine state-of-the art language and speech technology with world-leading expertise and facilities on (psycho-)linguistic research.
.
It concerns ESR8 in H2020-MSCA-ITN-2017 project TAPAS: Training network on Automatic processing of PAthological Speech [see also http://hstrik.ruhosting.nl/esr8-intelligibility/].
You will work together with 14 other PhDs and experts from European universities and companies that carry out research on this topic.
For young researchers an MSCA is an exiting opportunity for interdisciplinary training and research, and an excellent way to start your career.