Semantic taxonomy

My focus of late has been on the things I know and am closest to. As I analyze this, it is mostly because (not to beat a dead horse) I am a job seeker and when job hunting we push our strong suit. If push came to shove and I had to describe myself I would angle my classification of self as a soldier of study--I fight to keep learning. My biggest joy is comparing what I've learned today to what my kids have learned. When we get our first chance to speak at any given time, I ask "What did you learn today?" and often I hear back "Nothing; today was a review day." "Nothing?!?" I will challenge them "...EVERYONE learns ALL the time!" After a good eye roll and a deep sigh they try and turn the question around and I am always ready for them.

It was with the realization that I have been so outwardly focused on selling my wares and playing up my strong suit that I felt as though I had ceased my quest for fresh knowledge. Meditating on this I thought long and hard about what keywords had been popping up in disparate conversations. What memes had been at play that I had not been attuned to?

The answer to this soul-searching question is "Semantics" and "Taxonomy." It was that simple. I had been having conversations with former co-workers, sparring partners, executives, and college classmates and these words were popping up in conversations independently. Now I am supposedly a smart guy, although sometimes I feel like arguing that point, but I discovered that my understanding of those two words was only cursory.

I decided to delve into it further. So as to only plagiarize from my fellow soldiers, I won't quote a commercial dictionary (although I did consult a few), but I'll pull definitions from my favorite meme repository: Wikipedia.

Semantics:
Semantics is the study of meaning. The word "semantics" itself denotes a range of ideas, from the popular to the highly technical. It is often used in ordinary language to denote a problem of understanding that comes down to word selection or connotation.

Taxonomy:
Taxonomy is the practice and science of classification. In addition, the word is also used as a count noun: a taxonomy, or taxonomic scheme, is a particular classification ("the taxonomy of ..."), arranged in a hierarchical structure.

A search on Semantic Taxonomy did yield a few interesting hits. My first result yielded an interesting abstract on improved web search via A Personalizable Agent for Semantic Taxonomy-Based Web Search. The authors combine their expertise in business, computer science, and engineering to address "the problem of specifying Web searches and retrieving, filtering, and rating Web pages so as to improve the relevance and quality of hits, based on the user’s search intent and preferences." This article was the first place I found use of the term "meta-knowledge." The goal is to determine the user's intent under various conditions and allow the user to construct a mechanism for derivation and instruct the search engine in the nuances of criteria for various search conditions. The study relies heavily on the user pre-anticipating search goals, but I would hypothesize that with enough empirical data, the system could start to denote trends and might be able to intuit valid and relevant results without as much training. This is a fascinating concept.

The second result in my search was geared towards linguistics. This abstract, Semantic Taxonomy Induction from Heterogenous Evidence, proposes "a novel algorithm for inducing semantic taxonomies. Previous algorithms for taxonomy induction have typically focused on independent classifiers for discovering new single relationships based on hand-constructed or automatically discovered textual patterns." In contrast to the first abstract's use of "meta-knowledge" this study instead cites "global knowledge." An interesting departure, far closer to my vision of intuited results. I admit that their pages of formulaic ponderance left me a bit cross-eyed, but they bring it home in terms of seeking applicable novel hyponyms for perceived hypernyms. While the study spends a lot of algebraic energy I couldn't fathom in my brief recognizance, their conclusions corroborate my impression of the first abstract: with enough empirical data and good formulation, semantic taxonomies will bring increasingly relevant information to the fingertips.

So my cursory understanding of these two memes and how they find themselves intertwined is now much deeper. Semantics, still something I know I will find myself arguing for the sake of precision in communication, is also crucial to context within data models. Taxonomies, which I had been using correctly but had not given much consideration to their application in search environments, now hold significantly greater depth, both in terms of data structure as well as organization.

I cannot wait for my kids to turn the question "What have you learned today?" around tonight--this one's a doozy! So, for the sake of your own quest, please allow me to ask: what have YOU learned today?