[Xerte] Re: Dictionary Web Service

[Patrick Lockley]

Is the answer an "oracle" system?

With an oracle system, you get a definition by contributing a definition.

So each toolkits has a glossary page - you can request a definition, but also, all your definitions of new words are added to the system dynamically as you create the template?

To help seed the system - advise on other services as a starting point

Could easily support multiple definitions

-----Original Message-----
From: xerte-bounces at lists.nottingham.ac.uk [mailto:xerte-bounces at lists.nottingham.ac.uk] On Behalf Of Julian Tenney
Sent: 04 February 2011 10:02
To: Xerte discussion list
Subject: [Xerte] Re: Dictionary Web Service

I don't think the goal ought to be to try and produce a glossary of everything. Relying on existing sources is problematic because often folk want a particular definition, to highlight a particular point, or to nuance the wording in a particular way. I'm thinking pretty lo-fi, a database table with maybe as little as two fields 'word' and 'defn' and some simple methods of adding words and finding defns. Then that content can live outside of the actual content that presents it, and that solves the problem of this stuff living inside pieces of content, and makes it reusable... 


-----Original Message-----
From: xerte-bounces at lists.nottingham.ac.uk [mailto:xerte-bounces at lists.nottingham.ac.uk] On Behalf Of Fred Riley
Sent: 03 February 2011 13:54
To: Xerte discussion list
Subject: [Xerte] Re: Dictionary Web Service

> Like I said, it sounds like one of those beautifully simple ideas that
> if no one else has done we ought to do. It would fit well with the
> whole xpert / open Nottingham agenda.

Perhaps, but it would be an enormous project to create a public 'glossary of everything'. There are tens of thousands of nouns in English alone, and many of them would have different definitions in different contexts (off the top of my head - silicon, logic, up, spin, chip). You'd need a massive database to store definitions and tags and relations. Not to mention the aggravation of contributors fighting over wording. Essentially, you'd be creating a new dictionary from scratch, but without the 'extra bits' and of relatively limited application compared to dictionaries.

Maybe an XPERT-style aggregation approach might be more productive, if you could find individual glossaries with APIs/feeds you could harvest. 

Worth a try, and a useful project I'm sure, but you might run into 'reinventing the wheel' criticism. 

Cheers

Fred

_______________________________________________
Xerte mailing list
Xerte at lists.nottingham.ac.uk
http://lists.nottingham.ac.uk/mailman/listinfo/xerte

_______________________________________________
Xerte mailing list
Xerte at lists.nottingham.ac.uk
http://lists.nottingham.ac.uk/mailman/listinfo/xerte