Artificial-Intelligence Professor Makes a Search App to Outsmart Siri

部署运行你感兴趣的模型镜像

Siri, Google Voice Actions and other mobile speech assistant tools have promise, but they often fail to do what we ask them. And so we find ourselves scrolling one-fingered through mountains of messages looking for the right one, or painstakingly copy-pasting addresses into mobile maps.

 

Nick Cassimatis

A cognitive-science professor named Nick Cassimatis thinks that all these big companies are doing natural language search wrong, and he and two grad students have built their own set of tools — and a start-up called SkyPhrase — to try to show them up.

 

SkyPhrase today searches Gmail, Twitter and Orbitz, through its own Web site, Chrome extension and iPhone app.

It can search through email for attachments with a query like “emails that Jane sent me during the holidays containing pictures,” do complex Twitter searches like “tweets about Mars from NASA during the last two days,” and find airline tickets with flexible dates like “flights from New York to Orlando leaving next week and returning in November.”

Cassimatis, who is a professor at Rensselaer Polytechnic Institute and got his PhD in artificial intelligence at the MIT Media Lab, says the SkyPhrase approach is more precise because it is more linguistically informed than others.

“We memorize the dictionary to read the Library of Congress,” he said. “Siri is trying to memorize the Library of Congress.”

SkyPhrase is built to understand complex queries and syntactical relationships, not just keywords and commands. It can interpret conjunctions, coordinative clauses and noun phrases.

But there’s still a long way to go. SkyPhrase’s example queries may well stump Siri — as they did when I tried them — but it’s hard to even know what these services are capable of, much less dream up relevant queries in the moment. And the SkyPhrase site and app look more like a tech demo than a polished service.

The company — which has raised some seed funding and also accepts PayPal donations on its site — sells its Gmail search app for the iPhone for 99 cents.

The app does not search within the contents of emails or users’ contact lists, but understands subject lines, sender names, times and attachment types. Users can buy a paid subscription for speech recognition licensed from iSpeech for older iPhones, or just use the built-in speech recognition on newer ones. A free mobile Web version is coming soon, Cassimatis said.

What might be better than a specialized app — and perhaps more fitting, considering the SkyPhrase team’s expertise — is for Siri, Google and other people’s systems to get smarter using these linguistic search techniques. To that end, Cassimatis said he also plans to license a SkyPhrase API.

 

http://allthingsd.com/20121108/artificial-intelligence-professor-makes-a-search-app-to-outsmart-siri/

 

您可能感兴趣的与本文相关的镜像

Python3.10

Python3.10

Conda
Python

Python 是一种高级、解释型、通用的编程语言,以其简洁易读的语法而闻名,适用于广泛的应用,包括Web开发、数据分析、人工智能和自动化脚本

内容概要:本文介绍了一个基于Matlab的综合能源系统优化调度仿真资源,重点实现了含光热电站、有机朗肯循环(ORC)和电含光热电站、有机有机朗肯循环、P2G的综合能源优化调度(Matlab代码实现)转气(P2G)技术的冷、热、电多能互补系统的优化调度模型。该模型充分考虑多种能源形式的协同转换与利用,通过Matlab代码构建系统架构、设定约束条件并求解优化目标,旨在提升综合能源系统的运行效率与经济性,同时兼顾灵活性供需不确定性下的储能优化配置问题。文中还提到了相关仿真技术支持,如YALMIP工具包的应用,适用于复杂能源系统的建模与求解。; 适合人群:具备一定Matlab编程基础和能源系统背景知识的科研人员、研究生及工程技术人员,尤其适合从事综合能源系统、可再生能源利用、电力系统优化等方向的研究者。; 使用场景及目标:①研究含光热、ORC和P2G的多能系统协调调度机制;②开展考虑不确定性的储能优化配置与经济调度仿真;③学习Matlab在能源系统优化中的建模与求解方法,复现高水平论文(如EI期刊)中的算法案例。; 阅读建议:建议读者结合文档提供的网盘资源,下载完整代码和案例文件,按照目录顺序逐步学习,重点关注模型构建逻辑、约束设置与求解器调用方式,并通过修改参数进行仿真实验,加深对综合能源系统优化调度的理解。
评论
添加红包

请填写红包祝福语或标题

红包个数最小为10个

红包金额最低5元

当前余额3.43前往充值 >
需支付:10.00
成就一亿技术人!
领取后你会自动成为博主和红包主的粉丝 规则
hope_wisdom
发出的红包
实付
使用余额支付
点击重新获取
扫码支付
钱包余额 0

抵扣说明:

1.余额是钱包充值的虚拟货币,按照1:1的比例进行支付金额的抵扣。
2.余额无法直接购买下载,可以购买VIP、付费专栏及课程。

余额充值