TE
TechEcho
Home24h TopNewestBestAskShowJobs
GitHubTwitter
Home

TechEcho

A tech news platform built with Next.js, providing global tech news and discussions.

GitHubTwitter

Home

HomeNewestBestAskShowJobs

Resources

HackerNews APIOriginal HackerNewsNext.js

© 2025 TechEcho. All rights reserved.

Do you find WolframAlpha inflexible and unhelpful re: question phrasing?

3 pointsby CalmQuietalmost 16 years ago
My experience with WA has been disappointing. Although it sometimes shares a lot of <i>interesting</i> info, I experience little or no improvement for "ask Jeeves" sort of questions. It seems that there'll be quite a learning curve, and I am prone to turn elsewhere. How much <i>more likely</i> to be turned off is the (less-techie) public in general.<p>Here is the feedback I sent to WA (who have not responded to my earlier concerns). Perhaps HNers will have thoughts:<p>&#62;When you aren't "sure what to do with your(my) input"... WA certainly is <i>not</i> helpful in designing a search.<p>&#62;Today, for example, the news is hot with Dell's drop in sales. So I thought WolframAlpha might have links to stories comparing Dell with other manufacturers 2009 sales. Nope.<p>&#62;I entered into WA: computer server sales 2009<p>&#62;Result: <i>nothing</i> "Wolfram|Alpha isn't sure what to do with your input."<p>&#62;I <i>thought</i> that was a pretty specific, objective request (per your "tips for good results").<p>&#62;I would <i>like</i> to like WA... and have an alternative to Goog &#38; MS. Are your marketers &#38; user-interface people listening?<p>[/rant]

1 comment

ErrantXalmost 16 years ago
Such data is VERY new and so the chances of WA having it are low (the latest data I have seen is 08)<p>Dont forget WA isnt really into "popular" data - searching for stuff relevant to whats happening "now" wont find you anything productive. It's a computational engine and statistical data engine.<p>Also dont forget they only have 1000 data entry "experts".Only a teeny tiny percentage of useful data has been hand picked for inclusion yet.<p>Plus the responses are mostly fairly strongly hardcoded and the AI is very bad at recognising your questions unless it is a very mathematical term (basically it appears to try to apply math to the phrase, and then falls back to hard coded, keyword based searches).
评论 #632001 未加载