Скачать или смотреть Topic Modelling on G2 Customer Reviews with LLMs (Full guide)

🚀 Dive into the world of advanced review analysis with our latest video! We're tapping into the potential of OpenAI's LLMs to transform the way you understand customer feedback. Whether you're sifting through G2 reviews, support tickets, product ideas, or even YouTube comments, our step-by-step guide has you covered.

📊 Designed for Everyone

This video is perfect for data scientists, product managers, or anyone intrigued by AI's role in review analysis. It's tailored to equip you with the skills needed to effectively harness the power of customer opinions.

🔍 What You'll Get From the video:

• Topic Modelling Techniques: Uncover the hidden themes in customer feedback. See how LLMs like ChatGPT enable new forms of analysis.
• Practical Walkthrough: Follow along with my high-level as well as detailed code walkthrough and apply these skills in real-time.

💻 Full Code Included: Don't just watch – implement! We provide you with the full code to try out these techniques yourself.

Github Repo: https://github.com/balmasi/g2_reviews...

Want to leverage AI for your business? Book a discovery call: https://cal.com/borna/discovery

00:00 - Teaser of what we'll be building
00:42 - Concept: Overview of application of these techniques
01:43 - Concept: Downloading review data with Apify
02:10 - Concept: Splitting reviews
03:03 - Concept: Categorizing sentiment (Likes/Dislikes)
04:20 - Concept: Embeddings
06:19 - Concept: Clustering
07:00 - Concept: Naming clusters with LLMs
08:16 - Concept: Visualizing the reviews
09:11 - Web app demo
12:30 - Code: Downloading the project
13:39 - Customizing the G2 Data - Getting started with Apify
14:52 - Running the G2 scraper
17:27 - Code: Preprocessing the data
18:10 - Code: Splitting the reviews to sentences
18:37 - Code: Embedding the sentences
19:07 - Code: Clustering the embeddings
19:50 - Code: Getting LLMs to extract the topic of each cluster
22:44 - Code: Dimensionality reduction with UMAP for visualization
23:38 - Closing comments

Connect on LinkedIn:   / balmasi