Logo Utrecht University

Digital Humanities Benelux Conference 2017

Oral History and Speech Technology

Time & Room

Date & Time Monday July, 3 – 09:30 – 12:30
Room to be announced
Signing up Send an e-mail to DHBenelux (attn. Arjan van Hessen)

Goal and target audience

The objective of this workshop is to showcase to researchers in the field of media studies, Oral History and social scientists working with spoken narratives how the use of language and speech technology (HLT) can facilitate their research in their daily practice.
We will demonstrate various types of (partly open source) HLT-software, and will give an overview of the technologies that become available in the near future. Part of the workshop will consist of “hands-on-experience” where participants can use their own data and “our” tools.

Setup

In the workshop, we will address the following issues:

  1. Digitalisation:
    How to digitize analogue recordings (such as (cassette) tapes). What are the important issues, which software is available (and useful), what are the pitfalls and much more that is important in this often-crucial step from analogue to digital.
  2. Speech Recognition:
    Which digital tools are available to effciently make your own transcriptions? We will discuss the advantages and disadvantages of different types of Automatic Speech Recognition (ASR), of re-speaking and audio-text alignment. How to make a basic transcription with ASR that can be manually improved with the help of others.
  3. Forced Alignment:
    Forced Alignment (FA, the process to align audio with written text) can be done in most languages. We will show the use of WebMAUS for the forced alignment. Visitors are invited to bring their own digital text and audio.
  4. Transcription (Correction):
    In the workshop, we will show you software for making transcriptions and for correcting transcriptions that resulted from ASR.
  5. Metadata:
    Which formats are available and how to choose “the right one” for your own OH-projects?

Additional information

The workshop is an extension of the CLARIN-EU workshop (Arezzo, May 2017). Please have a look at this website for additional information about the OH-Transcription chain. A secundary objective of this DHBenelux workshop is to obtain feedback from OH-scholars in order to improve a proposal for a small European (CLARIN-wide) OH-infrastructure where existing and new technology can be used for OH-research.

Number of participants

We aim at a group of 10 – 15 participants.

Organisers

  • Dr. Arjan van Hessen (University of Twente)
  • Dr. Henk van den Heuvel (Radboud University Nijmegen)
  • Dr. Stef Scagliola (Université du Luxembourg)

Contact

For more information, please e-mail: Arjan van Hessen