Make your LearnDash video transcripts searchable

You can now autogenerate video transcripts for your LearnDash course videos using a free Python library called Whisper.

It is possible to take a folder of video files, and write some Python code to batch convert them into the corresponding mp3 audio file using a tool like ffmpeg, and then transcribe the audio file using Whisper.

There are many benefits of adding audio transcripts for your videos. But sometimes people hesitate because the cost of manual transcription is quite high, while automated transcription is usually not accurate enough. And the accuracy of automated transcription is usually lesser for domain-specific technical jargon, which are often the most important words in a video lesson. And things were usually worse for non-native English speakers who may have a noticeable accent.

Recent improvements due to Machine Learning

What changed in late 2022/early 2023 is that, because of advances in Machine Learning, we finally got accuracy rates high enough that

a) most of the technical jargon is now accurately recognized

b) it is almost independent of the speaker’s accent

On top of that, you can also format and improve the transcript using a Large Language Model like ChatGPT, and usually fix the few formatting and typos which still remain, and produce a very professional transcript which makes it much easier for your students.

Benefits of adding video transcripts

Here are some of the benefits of producing these automated transcripts:

Make it easy for students to navigate your videos

Obviously adding a transcript makes it much easier for your students to follow the video lessons. They can skim the transcript to get an idea what is covered in the video lesson.

And they can also do a CTRL + F to find specific words once they are on a specific lesson page, and then jump to that point in the video.

Create subheadings

Once you generate a transcript, you can use the GPT API to autogenerate subheadings (chapters) for your videos.

This makes it even easier for students to navigate your videos, as you can see from the example below.

Add chapters into the video itself

If you have suitable chapters, you can also add them into the video itself as you can see in the picture above.

For example, YouTube supports adding video chapters.

If you use video hosting like bunny.net, which is probably the best video hosting for LearnDash course creators, it also supports adding chapters into their videos. In fact, you can even automate the addition of these chapters using their API.

Search across video lessons

Another big advantage of doing this is that it makes your entire video catalog searchable, including the words spoken in your video lessons.

It will be possible for the students who have already purchased your courses to simply do a search across your entire website, find the lesson which discusses the topic they want to learn, and jump right into that lesson.

Create course previews for video lessons

By writing some Python code, you can extract the first few words of each lesson and display it as a course preview in LearnDash. (I am sure this is also possible for other LMS, but I am most familiar with LearnDash and I have already implemented it).

But this is not possible if you have just a video lesson, but no transcript.

However, once you do add a transcript, you can just use the first few words from the video transcript itself as the preview.

About this website

BotFlo1 was created by Aravind Mohanoor as a website which provided training and tools for non-programmers who were2 building Dialogflow chatbots.

This website has now expanded into other topics in Natural Language Processing, including the recent Large Language Models (GPT etc.) with a special focus on helping non-programmers identify and use the right tool for their specific NLP task. 

For example, when not to use GPT

1 BotFlo was previously called MiningBusinessData. That is why you see that name in many videos

2 And still are building Dialogflow chatbots. Dialogflow ES first evolved into Dialogflow CX, and Dialogflow CX itself evolved to add Generative AI features in mid-2023