Can an algorithm (or thousands of them) ever be able to mimic the complexity of human motivation? That is the gist of semantic search, which search engine companies like Google attempt to perfect. Semantic search is the idea that a searcher types “apple” into the Google search box, and based on history of searches, synonyms of the keyword, and keyword matching, the Google algorithms will “know” whether I meant the apple the fruit or Apple the company.
I read a blog post recently how Google is partnering with a semantic search engine company called Oingo, which has made great strides toward improving semantic search results. According to Oingo, “By searching on meanings, instead of just words, our search eliminates this “randomness” of results.” Oingo uses synonyms and alternative suggestions to help direct the searcher toward relevant content.
Semantic search is ingenious. It combines linguistics, cognitive psychology, philosophy, and computer engineering. I think it’s a fascinating attempt to create a “humanized” search mechanism without being human. I also think it will never be possible to perfect this science. Humans can barely understand and predict human motivation let alone using that limited knowledge to program an algorithm to do it for them. We shall see how long it will take semantic search engineers to realize that humans are innately unable to step outside their own brains, not semantically, psychologically, programmatically, or even metaphysically.
