Join The New York Semantic Web Meetup Group

You'll get invited to our Meetups as soon as they're scheduled!

The New York Semantic Web Meetup Group Message Board › reasoning vs. inferencing

reasoning vs. inferencing

Richard Creamer
Posted Jun 28, 2009 4:19 PM
rtc1
Folsom, CA
Post #: 2
Send an Email Post a Greeting
Hi,

I have a basic question:

Is there a difference between reasoning and inferencing within the context of the Semantic Web?

Thanks,

Rick
Marco Neumann
Posted Jun 28, 2009 4:59 PM
user 2316446
Group Organizer
New York, NY
Post #: 28
Send an Email You are unable to greet this member
The terms are used interchangeably. Wikipedia has a nice page on Inferencing
http://en.wikipedia.o...

On the Semantic Web you reason to infer new facts or perform consistency checks with a Semantic reasoner.
http://en.wikipedia.o...
George
Posted Jul 31, 2009 12:02 PM
user 5879939
Newtown, PA
Post #: 1
Send an Email Post a Greeting
I think there is a subtle difference between reasoning and inference, outside of the Semantic Web.

Reasoning can be based on using one or more sets of 'consistent' logical frameworks and inference is typically based on one logical framework.

One of the best quotes I ever head about reasoning was this from Sherlock Holmes (okay Sir Arthur Conan Doyle, Sr.):

“Once you eliminate the impossible, whatever remains, no matter how improbable, must be the truth.”

This is a type of reasoning that is tough to create as an inference, but makes perfect sense as a reasoning tool.

From an inference standpoint, the above quote makes no sense because you can only infer what is possible given the logical framework you are using.

But back in the Semantic Web world, we are only working within a logical framework defined by the data and so the two terms can be used interchangeably.

Of course, this is my view and your mileage may vary...

Aileen Gemma Fedul...
Posted Aug 31, 2009 10:17 PM
user 10289793
Staten Island, NY
Post #: 1
Send an Email Post a Greeting
Hi George and everyone,

Thanks for sharing the quote.

Actually this discussion prompts the crux of a different issue to me...

In one sense the semantic web is a graph that moves back and forth with different gears and different kinds of relationships. So from that point of view you are looking at a logical framework served by the data- thus inference and reasoning can be used interchangeably.

But then you have this small, but consistent problem known as people who read and understand and think of those terms in different ways.

We have the nuanced understanding that a writer, a poet, an engineer, a grade school teacher, a teenager and a doctor all bring to the same set of words or phrases. Layer that over with several differences when you get to different languages and to different generations. For example something you could say quite precisely in Japanese or German but not American english, and something I might say as a woman in her mid thirties, "Let me ring you" versus what my father (his mid seventies) would say- "Let me give you a buzz later..."

What I am getting at is part of this disconnect (does it matter if it is inference or reasoning) may be in how we as people understand and process language and we must be sensitive to that in how we build these relationships.

So it's not merely a matter of what the graph is or how it functions. If only we could all get our words to mean the same thing, to borrow from Lewis Carroll...`When I make a word do a lot of work like that,' said Humpty Dumpty, `I always pay it extra.'

I am very eager and curious to know what others think...

With kindest regards,
Aileen

Powered by mvnForum