Tech

Providing Attendees with Transcripts for Online Events

Written by The Meeting Pool

What considerations do planners need to take in order to provide attendees with accurate and accessible transcripts of online events? A recent conversation on Catalyst discusses the details.

Author:
Convene Editors       

Buffer

woman transcriptionist in headphones illustration

Offering transcripts of online events is “a good thing to do and one can argue it’s fast becoming essential,” said one member of PCMA’s Catalyst community.

PCMA’s Catalyst community offers members a platform to ask each other questions, share ideas, or, as the website says, “communicate and collaborate.” Here’s a sampling from a recent Catalyst discussion.


“Is anyone providing transcripts with the recordings of their online events?” Britta Newell, director, events and experience, Institute for Functional Medicine, asked on the Catalyst forum. “It seems like the right thing to do to increase accessibility, but I’m also concerned about intellectual property issues. I’m trying to get a feeling for what users expect from event tech these days.”


An interesting question. We deliver health-care events around the globe for associations, pharmaceutical companies, and institutes like yours. We see a broad range of examples.

It’s certainly becoming more common, but usually there are a few decision steps to go through. There are so many platforms available that any IT-savvy viewers can download a recording and watch it through and create their own transcript free or at relatively low cost. So, if the recording is already in the public domain, adding a transcript is (in my opinion) unlikely to cause you a problem. Not having it accurate might! If you make the decision to offer recordings, I imagine that you have the speaker permission to do so. Let’s assume that is the case.

Next decision point is accuracy. We tend to find that at best you should expect AI-driven transcripts to generate content [that is] 90-percent accurate. If you want to increase accuracy, you could preload key medical terms and a glossary of common terms and this might get it up toward 95-percent accuracy. To get it 100-percent accurate, you need to involve a human! Where it falls over is accent (I’m Scottish, and some AI systems don’t cope well with a strong Scottish accent) and where the presentation has similar but different acronyms related to a topic. If there is a need for compliance reasons or accuracy of content, you may again need to use the transcript to allow post-event editing of the recording. Therefore, accuracy is key.

The other point that you mention is accessibility. Over the last 12 months, we have produced events in English (and other languages) that have reached delegates across 149 countries across multiple languages. I find that in many cases, we have a bunch of delegates who do not speak English as a first language therefore offering real-time transcripts and post-event transcripts not only increases accessibility, but significantly improves understanding and knowledge transfer. It’s a good thing to do and one can argue it’s fast becoming essential.

Finally, we recently delivered a global event where we had to take the original recording and then hard code the transcript in nine different languages for different regions of the world — simplified Chinese, Korean, Portuguese, etc. To achieve this, we had to get the English transcript right to have any chance of translating that transcript again through to different languages and finally getting it 100-percent accurate with human intervention! Delegates could listen in English and watch the transcript in their chosen language.

— Leslie Robertson, CEO & Founder, Open Audience


Read more Catalyst questions and comments

X