First, there are two classes of “Answer Engines”; computer-based (like Wolfram Alpha) and human-powered (like Quora.)
Each has their own benefits and limitations.
When you ask a computer-based system, it is using an algorithm to first translate what your question means, then has to look into its database of possible answers to find a match. It will give you the best possible match it can find based on what it has available.
If you search for a term it doesn’t know, generally it will translate it to the closest thing that it does know, assuming the likelihood is that you typed it incorrectly or that you would rather see AN answer rather than none.
There will be a completely non-emotional, fact-based response given to your query.
When you ask a human-based system, you will have a real person reading and answering your question. If you type something incorrectly, the human being will most likely mentally correct that mistake in answering.
They will be able to interpret nuance in the question. (Knowledge-engine vs Answer-Engine for example.)
And they will be able to give you an experiential response, that is emotion-based and not merely fact-based.
Some sites, like Yahoo have made a name for themselves for both the egalitarian quality of the questions and their resultant answers.
TL;DR: Humans will give you their personal “best fit” answer. Computers will give you the “whole” one – either wholly correct or wholly incorrect.
Originally Posted: https://www.quora.com/What-are-the-limitations-and-benefits-of-answer-engines-today
Originally Posted On: 2015-04-07