Lynne Bowker: Machine Translation Literacy for the Scholarly Community

TL;DR

• Machine translation has evolved, presenting ethical concerns and challenges requiring key insights for informed usage.

Transcript

so I'm sure you all know the history of machine translation I'm not going to spend a long time on it but essentially we are in a state of a kind of paradigm shift at the moment if you think back to the early days of machine translation the sort of post Second World War period the general approach to machine translation was to sort of try and make c... Read More

Key Insights

• 🎰 Paradigm shift: Machine translation has transitioned from rule-based systems to neural machine translation, leveraging AI and machine learning.
• 😒 Challenges: Easy accessibility and unthinking use, fluent but inaccurate translations, and disintermediation threat in the translation profession.
• 🤨 Privacy concerns: Machine translation platforms store user data for training, raising ethical issues regarding data privacy and usage.
• 🎰 Output improvement: Simplifying language structures, using plain language, and ensuring terminological consistency can enhance machine translation output quality.
• 🎰 Machine bias: Algorithmic biases can impact machine translation accuracy, necessitating user awareness and verification of machine-translated content.
• 🎰 Workshop outcomes: Initial workshops on machine translation literacy received positive feedback, suggesting interest in tailored workshops for specific language pairs or tasks.
• 🦻 Potential applications: Machine translation literacy workshops could benefit school and public libraries, aiding newcomers and immigrants with diverse language backgrounds.

Questions & Answers

Q: How has machine translation evolved over time?

Machine translation has shifted from rule-based systems to neural machine translation, utilizing AI and machine learning for better quality output.

Q: What are the challenges associated with machine translation usability?

Challenges include easy access leading to unthinking use, overly fluent but inaccurate translations, and the threat of disintermediation affecting translation professions.

Q: What concerns arise regarding machine translation privacy?

Utilizing free online machine translation services raises privacy concerns as these platforms retain and potentially use user data for training machine learning algorithms.

Q: How can users improve machine translation output quality?

By simplifying language structures, using plain language, maintaining terminological consistency, and being mindful of input quality, users can enhance machine translation output effectiveness.

Summary & Key Takeaways

• Machine translation has undergone a paradigm shift from rule-based systems to neural machine translation.

• Challenges include easy accessibility leading to unthinking use, fluent but inaccurate translations, and the threat of disintermediation in the translation profession.

• Promoting machine translation literacy involves understanding different tasks, privacy concerns, algorithmic biases, and improving output through input modifications.