Re: Mathematica syntax in W|A
- To: mathgroup at smc.vnet.net
- Subject: [mg102127] Re: Mathematica syntax in W|A
- From: Erik Max Francis <max at alcyone.com>
- Date: Thu, 30 Jul 2009 05:31:55 -0400 (EDT)
- References: <h4p3d7$imm$1@smc.vnet.net>
Filippo Miatto wrote: > Dear all, > I was wondering to what extent one can use Mathematica syntax for > WolframAlpha queries. > I mean, is there a more complex way to write queries that could be > used for complex searches? To access the data in a more effective way? > Stephen Wolfram in a webcast showed some hints (for example to add > ip=FR or something like that so that WA would interpret the query as > it was coming from a european country and use the metric system of > measures), I wonder if there is more. Nope, from what I've seen the syntax is not at all Mathematica-like. Which kind of makes sense, since if it could understand Mathematica proper, there'd be little reason to buy a Mathematica license. Obviously that isn't the goal of Wolfram|Alpha, though it's still not clear to me what long-term business prospects it has or is supposed to have. The natural language parser really leaves a lot to be desired. I made a suggestion (submitted the usual way through the online form, undoubtedly lost in the noise) that they do something reasonable, like give special dispensation to obvious clue words, like _and_, _or_, _of_, _in_, etc. for guessing separate queries, conversions, etc. This could even form a new (but simple) language that could easily be used by even non-technical endusers. It doesn't seem like there's going to be any substantial improvement if they're going to keep doing statistical matching (or whatever it is they're doing) by basically randomly guessing what queries mean. -- Erik Max Francis && max at alcyone.com && http://www.alcyone.com/max/ San Jose, CA, USA && 37 18 N 121 57 W && AIM/Y!M/Skype erikmaxfrancis Nothing is so good it lasts eternally -- Florence, _Chess_