Hi everyone, so I was looking into how to get started with creating a matching algorithm for a project at work. I do realize this sounds like how to get to the moon in 5 easy steps, but I was just wondering if anyone could point me in the right direction to learn and do some research regarding the topic.
I consider myself to be a front-end developer with some experience in full-stack development, but when it comes to applying a matching algorithm to a functional community platform to curate content according to the user’s interests I am a bit lost. The last time I did algorithms was at university during my computer science studies.
I am willing to put in the time to educate myself, but currently using Google I am unable to find a jumpoff point.
Thanks it will be mostly matching user-generated content (status updates / shared images / links) to users with similar interests and then also system-generated content (ads / articles / product placings) to potentially interested users.
As a college-educated professional you should understand the necessity and methodology of analysis. You begin by stating the requirements in a user-oriented manner then create some technical designs. You can’t find an appropriate algorithm if you do not know what is appropriate.
What are status updates? Do you mean text? Does that require parsing of “Natural” (people) language? Look for Natural language. It is highly complex. What do you mean by shared images? Do you mean an URL? Is it just a matter of comparing complete URLs? Same for links.
Also, for matching links, instead of matching the URLs you could try matching metadata; I am not sure of the terminology, but probably SEO data. That would be more complicated than you might think since you should match synonyms too, like Google does. For example most people would consider Kleenex to be a synonym for tissue. And bracelet might be a synonym for wrist band.