A few days ago, I was searching for variations on the phrase "weigh search results." ^1 I was thinking about how I could alter the information retrieval logic used in my LLM chatbot ^2 to prefer content that was published more recently. I had two factors that I wanted to weigh: document similarity and document recency. While trying to figure out the best way to do this -- a topic I am still evaluating -- I found that what I had written on weighing search results had appeared at the top of my search term on Google.
Interestingly, the general principles of what I describe in the post -- assigning different weights to factors chosen for ranking -- holds true and relevant, but the means by which I could weigh factors with data on entirely different scales such as that with which I was dealing for my LLM chatbot was not addressed by the article. Indeed, the phrase "rank search results" (as "rank" is part of the search engine parlance) may yield results to further my research on this topic. Thus, my search goes on, knowing that I have been here before in some ways, documented what I learned then, and will likely write about the solution to this problem when I find one that works.
We need a term for seeing something you have written rank highly on a search engine, if one does not already exist. "First page blues" could refer to seeing your content in a search engine that doesn't exactly address your query; "first page joy" could refer to the excitement you feel when stumbling upon something you have written that helps solve a problem. I'm making up these terms. Perhaps there are already ones out there that could be used instead?
Regarding my technical query, if you know how to weigh semantic similarity and document recency or want to brainstorm on the topic, let me know!