Navigation auf uzh.ch
Navigation auf uzh.ch
Blind and visually impaired, deaf and hearing-impaired, cognitively, motor-impaired, and persons with speech and language disorders face many barriers in their everyday lives, often related to language. This course provides an overview of common barriers and introduces artificial intelligence approaches developed to reduce some of these barriers. Specifically, the course deals with tasks such as sign language recognition, translation, and production; intralingual subtitling; audio description; diagnostics of speech and language disorders; automatic text simplification; and speech recognition and synthesis as part of Augmentative and Alternative Communication (AAC) and Ambient Assisted Living (AAL). A focus is on research approaches; transversal topics are those of multimodality and ethics. Students will gain hands-on practice applying some of the approaches as part of the exercises accompanying the course.
Blind and visually impaired, deaf and hearing-impaired, cognitively, motor-impaired, and persons with speech and language disorders face many barriers in their everyday lives, often related to access to information and communication. This course provides an overview of common barriers and introduces language-based assistive technologies and e-accessibility measures. Specifically, the course deals with guidelines for creating accessible electronic documents (e.g., Web pages), with screen readers, Braille, operation of mobile devices, sign language, tactile signing, Augmentative and Alternative Communication (AAC), and ethics. Students will gain hands-on practice using various tools as part of the exercises accompanying the course. This course is complemented by a course on Master's level with a focus on deep learning approaches.
Um die Digitalisierung als Ganzes zu begreifen und um erfolgreich studieren zu können, muss man verstehen, was «Digitalisierung» ist. Deshalb braucht es eine digitale Grundkompetenz. Ziel der von der Digital Society Initiative der UZH entwickelten Online-Vorlesung «Studium Digitale» ist es, eine erste Einführung in dieses weite und komplexe Thema zu geben. Didaktisches Konzept: Online-Kurs (auf Deutsch) bestehend aus Videolektionen, Übungen und Selbstlerntests. Der Kurs wird auf OLAT durchgeführt. Die inhaltliche Kursbetreuung wird von den Dozierenden (oder ggf. ausgewählten Mitarbeitenden) geleistet, die generelle Maintenance und administrative Betreuung liegt bei der DSI.
The aim of this course is to deepen programming skills in Python and to introduce concepts of modern software engineering (version control, project management, testing). Relevant topics and techniques are conveyed through selected applications from the fields of computational linguistics and natural language processing.
While the topic of ethics in natural language processing is receiving more and more attention, ethical aspects pertaining to certain sub- groups of users are still underresearched. This is especially true for the area of assistive technologies in the context of persons with disabilities. The aim of this seminar is to gain an understanding of the relevance of ethical aspects in the aforementioned area, from bias in data to participatory/inclusive research design. We will also touch on the aspect of data protection. Seminar contributions can be of practical or theoretical nature.
Most research on Language Technology focuses on 20 languages and relies on large collections of data. In this course, we discuss approaches to Language Technology for less studied and lower resourced languages (LRL). We investigate strategies to deal with data scarcity for specific Language Technology tasks (e.g. Part-Of-Speech-Tagging, Named-Entity Recognition and Machine Translation). We also look at the impact Language Technology has on Low-Resource language communities.
Language technology has contributions to make to diagnostics, therapy, and instruction of persons with disabilities and special educational needs. For example, speech recognition can aid in automatically transcribing spontaneous speech samples. Text analysis and classification are capable of supporting the diagnosis of language disorders. As part of therapy/intervention, embodied conversational agents are used to increase pragmatic skills in persons with autism spectrum disorder, while speech recognition is used in dysarthria therapy. The aim of this seminar is to gain an overview of current approaches to (language-)technology-supported diagnostics, therapy, and instruction in the context of special education. The seminar will be complemented with guest talks from fields such as special education and cognitive science and as such will take an interdisciplinary stance.
