Keyword Extraction using ChatGPT

We’ve got summaries of the podcast.

BTW. it cost me $35 to retrieve all of it, but you can get it for free here to continue following along this tutorial.

The giveaway I linked above contains summaries with keywords.

The keywords let you search easily through documents in apps like Obsidian.

Today we will see how to extract these keywords from summaries using Langchain.

This is the prompt that lets us achieve this:

You are a highly-efficient expert of knowledge databases

Your task is to assign high-level keyword categories to summaries of podcast segments you receive.

These keywrods will be used by users to search through podcast material. 
That means, the keyword categories you assign should be easy to search through and accurate.

As an input, you will receive a text with summary.

As an output, I want Python list of three to five elements with keyword categories (each category should consists of 1-2 words).
Output only the list, no other text nor description.

SUMMARY:
```
{text}
```

ASSIGNED KEYWORDS (Python list):

We used this prompt to iterate over all files with summaries and attached the keywords on top of the file. We also added an embed of a YouTube video, so we can view it from inside the note. The notebook that lets us achieve this can be found here.

This is what an example result looked like:

What’s even more cool - in Obsidian, we can search for relevant notes through hashtags (e.g. find notes that contain artificial_intelligence AND love tags):

This is the ninth day of the 30-day AI challenge.

Over the next month, I will be building the Lex Fridman AI engine with you!

If you're reading this, I assume you'd like to build things. If you stick to this newsletter you will have a running project after a month and know the necessary technology to build AI apps.

I've recently built PodcastGPT and want to share the process with the community. If you haven't seen the app yet, you can get access here: PodcastGPT

This is all for now! See you tomorrow.

Stay focused!

Luke