Chapter 13. Expressiveness and Content Dilemmas

I don’t know how much more expressive you can get than being a rock and roll singer.

Robert Plant

In this chapter, we deal with dilemmas about what should be included in a semantic model and what can (or should) be left out. As a model creator, this will help you achieve the right balance of expressivity and content that your model needs without wasting effort and resources. As a model user, on the other hand, you will be able to better understand the rationale behind certain expressiveness and content choices that different models make (e.g., not having complete lexicalizations of entities, or representing multiple truths) and adjust your expectations accordingly.

What Lexicalizations to Have?

Some 30 years ago, when I was a kid, I used to watch a sportscaster on a Greek TV station who would routinely refer to the English Premier League soccer clubs by their nicknames: “the Gunners” for Arsenal F.C., “the Citizens” for Manchester City, “the Magpies” for Newcastle United. I had no idea why he did that and, in the beginning, I was annoyed because I couldn’t understand him. After several viewings of his sportscasts, though, I also started calling these teams by their nicknames instead of their formal names, and I bet I wasn’t the only one to adopt these nicknames. Thirty years later, every time I build a semantic model and I need to provide lexicalizations for its entities, I can’t help but think: What would this sportscaster call ...

Get Semantic Modeling for Data now with O’Reilly online learning.

O’Reilly members experience live online training, plus books, videos, and digital content from 200+ publishers.