INTERACT – The International Network in Crisis Translation

Health and Well-being

INTERACT – The International Network in Crisis Translation

Project synopsis

Multilingual translation of crisis and disaster-related content is a necessity and a human right for health, safety and resilience. Crisis Translation is understood as the translation of written information from one linguistic and cultural system to another in the context of a crisis scenario, with a view to enabling affected communities and responders to be prepared for crises, improve resilience and reduce the loss of lives. Crisis communication must be multilingual and multilingual crisis communication is enabled through translation. Multilingual information access through translation addresses societal challenges such as social fairness and democratic access to essential information for all.  However, translation and interpreting have been sorely neglected in the field of emergency and disaster risk reduction. This project tackles that neglect through policy, training, technology and ethics-based research.

Cochrane.org
Translators without Borders
New Zealand Red Cross

The primary focus of INTERACT is on health-related crisis content. The main objectives of the project are:

  1. to make meaningful and effective contributions to knowledge, policies, expertise, training and technology that enable accurate and timely translation-enabled crisis communication, with a focus on health-related content;
  2. to improve outcomes for crisis-affected and at-risk communities by contributing to translation-enabled communication to, with and from those communities;
  3. to enhance human skills, competences and cross-sectoral collaboration across academic, humanitarian, and industrial sectors involved in crisis translation. This is achieved by focusing on five main topics: crisis communication policy, comprehension in limited English proficiency communities, crisis machine translation, citizen translator education, and ethics and through collaboration across academic, SME, NGO, and Multinational partners.

Each pillar within the project requires different research methods. We make use of the interdisciplinary skills and knowledge of the INTERACT team for each of these methods as follows:

  • Policy – systematic multilingual review using a human-rights based 4-A (Availability, Accessibility, Acceptability, Adaptability) framework of analysis.
  • Comprehension of Health Content – Survey instruments, followed by standard usability testing (Software Usability Scale) and traditional comprehension and recall measurements from cognitive science and reading studies.
  • Crisis Machine Translation – Neural approaches to artificial intelligence to discover best methods for creating machine translation systems for low-resource and ‘unknown’ languages. Standard automatic and human translation evaluation techniques.
  • Training Citizen Translators – Participatory Research Methods applied to cycles of pedagogical design and delivery/evaluation of content.
  • Ethics – Systematic review of content for ethical frameworks for policy, training, technology use; we are deploying a trust-based, virtue ethics approach.

The outputs are as follows:

  • Publicly available course content for Citizen Translators
    • Content for citizens and for students of translation has already been developed and delivered and we will continue to evaluate and evolve this over the duration of the project (to March 2020 and beyond).
  • Public reports on policy for national emergency response and how it takes multilingualism and the provision of translation into account; Identification of best practices and recommendations
    • First analysis of five national policies is complete and analysis of several other national policies is underway; This will feed into a public report by the end of the project (2020).
  • Preparation of journal articles on policy, ethics, training, simplification and technology is underway
  • Pilot Machine Translation engines that ‘pivot’ across three languages
    • Demonstrating current capability given low linguistic resources – end of project 2020.
  • Dissemination of findings through conferences, podcasts, Twitter, website and public dissemination events is underway. We have so far presented at five conferences since the start of the project (April 2017) and three more conferences are scheduled for 2018.

The outcomes are as follows:

  • Ensure that ethical use of translators and technologies is foregrounded in humanitarian response.
  • Ultimately, provide those affected by disasters content in a language they can understand in order to improve outcomes.
  • Societal Engagement
  • Health and Wellbeing
  • Professional Services
  • New Knowledge and
  • Human capacity
Higher Education Institute:

DCU

Funding Source:

EU H2020 Marie Curie RISE programme

Academic Contact:

Dr Sharon O’Brien,
School of Applied Language and Intercultural Studies
CTTS & ADAPT Research Centres