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.

Talk to Your Data: One Model, Any Relational Database

67 pointsby atrudeauover 4 years ago

7 comments

DevX101over 4 years ago
I&#x27;ve done some previous digging into natural language SQL queries -- there&#x27;s a good amount of research around this. But the error rate is always too high for useful production deployment and the systems I&#x27;ve looked at never handled ambiguity well. The target user for this is someone who knows nothing about SQL so ambiguity is guaranteed.<p>Most importantly, you can&#x27;t have the system confidently answer incorrectly.<p>I just tested this query to evaluate joins: &quot;What is the name of the series with the highest rating?&quot; The answer: 5.8 :&#x2F;<p>A typical user will be even less clear than this, asking the question as: &quot;What&#x27;s the top tv series?&quot;<p>The ideal system should be able to confirm with the user that 1) they&#x27;d actually like to evaluate tv series based on rating. And 2) it should be able to guess without user confirmation that I really want the &#x27;name&#x27; of the tv series.<p>It&#x27;s a hard problem but massively valuable if it can be solved.
评论 #24287789 未加载
评论 #24287590 未加载
评论 #24291201 未加载
mushufasaover 4 years ago
I tried &quot;What are the names of all the workers&quot; but it didn&#x27;t understand. So it doesn&#x27;t seem to take slightly-vague input. (The schema has a workers table with fields first name and lastname)<p>When I have to be precise, I&#x27;m better off using SQL directly.
thelazydogsbackover 4 years ago
IMHO, this type of transformation is best done via traditional symbolic methods that also provide explanations with support, alternate outputs based on context&#x2F;constraints, etc. It&#x27;s ok for a symbolic transformation system to have the <i>help</i> of connectionist black-boxes that act as oracles, but the entire <i>system</i> should not be a black box.
lmeyerovover 4 years ago
This has been one of the more compelling demo areas for GPT-3 as well (ex: <a href="https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=WlMHYEFt2uA" rel="nofollow">https:&#x2F;&#x2F;www.youtube.com&#x2F;watch?v=WlMHYEFt2uA</a>) -- are there any open AI efforts here?
zeropover 4 years ago
Does it work for browsing by dates.. real user may ask date range in various ways.. e.g. give my previous month&#x2F; a date&#x2F;quarter&#x2F;weeks etc.. the demo didn&#x27;t have dates in data I belive
评论 #24286668 未加载
ordxover 4 years ago
Looks like it expects user to know column names. Column names like total_price, fname, user_lname, etc. would make question answering a bit awkward.
homarpover 4 years ago
demo: <a href="https:&#x2F;&#x2F;naturalsql.com&#x2F;" rel="nofollow">https:&#x2F;&#x2F;naturalsql.com&#x2F;</a>
评论 #24286123 未加载
评论 #24286788 未加载