Don’t bother asking your computer “why” just yet

Picture source: screenshot of Google results.
Picture source: screenshot of Google results.

We’re pretty good at getting computers to answer certain kinds of questions.  A good example is what people in my field call factoid questions.  These are questions that have a clear answer, typically some sort of a noun, and usually pretty short.   They tend to start with words like who, what, when, or where.  What year was Mozart born?  Where was the first McDonald’s?  Who wrote Pride and Prejudice?  There’s been a ton of research on how to get computers to answer these kinds of questions, much of it organized by the National Institute of Standards and Technology’s annual Text Retrieval Conference.

Although we’re pretty good at getting computers to answer factoid questions, it’s still really difficult to get a computer to answer a why question.  Unlike factoid questions, whose answers are typically a short phrase, why questions are usually answered by an explanation or a procedure, and these are typically longer than a single phrase–Suzan Verberne, a researcher at Radboud University Nijmegen and an expert on why questions, found that answers to why questions are typically at least one full sentence in length, and can easily be as long as a paragraph.  We’re not nearly as good at finding these longer stretches of text as we are at finding those little “factoids.”

In English, the field of research that deals with getting computers to answer questions is called question-answering.  In French, it’s called questions-réponses.  Here is a verb that I learnt from the French Wikipedia page on question-answering:

  • se fonder sur qqch: to be based on something, to be built on something.

Les Systèmes de réponse à des questions explorent de nouvelles méthodes de recherche d’information exploitant des requêtes formulées à l’aide du langage naturel et non plus en se fondant uniquement sur des mots clés (comme c’est le cas avec les moteurs de recherches actuels).  “Question-answering systems explore new methods of information retrieval, using requests formulatd in natural language and not being based only on keywords (as is the case with current search engines).”

2 thoughts on “Don’t bother asking your computer “why” just yet”

  1. “why” is pretty sophisticated question, open to both objective and subjective interpretations. I don’t see computers dealing reliably with questions other than clear-cut “cause-effect” issues (e.g. scientific data)?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

Curative Power of Medical Data

JCDL 2020 Workshop on Biomedical Natural Language Processing


Criminal Curiosities


Biomedical natural language processing

Mostly Mammoths

but other things that fascinate me, too


Adventures in natural history collections

Our French Oasis


ACL 2017

PC Chairs Blog

Abby Mullen

A site about history and life

EFL Notes

Random commentary on teaching English as a foreign language

Natural Language Processing

Université Paris-Centrale, Spring 2017

Speak Out in Spanish!

living and loving language




Exploring and venting about quantitative issues

%d bloggers like this: