This is How I used the GPT o1 Reasoning Model to Build the TF-IDF SERP Analyzer Chrome Extension

I was thinking about how great it would be if I could see how similar the titles are to a search query on Google.

This can help explain if SERP can use new food and if it already accepts new food. Now there are myriad ways of measuring Similarity.

You can take the N-Grams Route too. I preferred taking the TF-IDF (term frequency-inverse document frequency) Route wherein SERP Titles would be vectorized & assigned Cosine Similarity Score based on the title vs the search query.

I explained GPT my idea & it flawlessly helped me execute my idea 💡

Now as I conduct a Google Search I can see it overlay similarity score right next to the SERP Titles. And this includes Sitelink extensions & PAA SERP titles too.

In the attached snapshot you can see how my Chrome extension shows the similarity score for the SERP Titles.