On recommending

for those WHO search

Written by Ian Sharp PhD on Jan. 30th, 2019

You are listening to The Good Doctor Sharp on DoctorsInTech.com

Everyone is looking for something. Whether we're actively typing into a search box or when we're just living life. To search is to have an intent. It's the nature of being alive and human. How do you recommend search results?

But first understand this... your searches are searching. It's not just you that's searching. What we're addressing now is the topic of WHO or rather WHAT is searching. And yes, your searches are searching.

Read my previous blog post "Which came first, the search query or the search result?" to better understand. You would have learned a little more about the nature of information itself, and I recommend that you read that blog post before reading on.

Your searches are searching because the overall concept of a search result document exists, in many forms, whether you acknowledge any of those forms or not. "Can you give me an example of that?", you might ask? Well, yes, I sure can and I'd be glad to. How about the 'category'.

If you're reading a recipe about making soup, you might contest that the entire recipe is about "cooking at home". So "cooking at home", could be the category of your reading material. I'll give you another example category. What about "soup"? The recipe is about making soup isn't it. Different documents can have different categories. A machine can tell you what the categories are, or you can come up with them yourself. It is also a large and exciting area of research.

That's what happens when you read articles about making soup and then you suddenly find ads asking you to try the new england clam chowder. Read my next blog, where I touch on another important topic; a topic that falls within the same category of big data.


Ian Sharp PhD

Need a recommendation engine for your business? Reach out and request your free case study today.


since 2014