最有看点的互联网金融门户

最有看点的互联网金融门户
国际资讯金融信息服务

用AI捕获市场信息变动 Amareos要抢投行的饭碗了

21世纪初,对冲基金的所有者们突然有了一个想法——金融市场受信息的影响,所以如果你能快速而透彻地读懂新闻,你就有机会成为市场中第一个做出反应的人。使用人工智能技术,他们设计出了可以实时分析突发新闻的软件,评估其含义,并在几秒钟内基于分析结果做出交易决策。

与对冲基金世界中的很多事物一样,这一人工智能产品一直是亿万富翁和专业投资者的专属工具。但一家名为Amareos的香港金融科技创业公司正计划将这类技术带给寻常百姓。

测量世界的温度

“我们在试图将大数据普及化”,Aamreos的联合创始人Philippe El-asmar表示。

Amareos达到了你对“大数据”的任何期待。这一服务可以分析超过100万个网站以及大量的社交媒体。使用由合作伙伴汤森路透及Market Psych开发的人工智能系统,Amareos可以测量互联网世界中特定话题的热度。

想要知道国际市场对美元的情绪?Amareos可以分析全世界对这一货币的讨论并向用户反馈一份详细的舆情分解——多少人对此表示愤怒,多少人充满信任,又有多少人感到厌恶。

1 第1张
今天互联网世界对美元的态度的图表

“我可以告诉你现在互联网对标普500有多少的快乐情绪,以及人们对人民币有什么感受”Philippe说,“一个由心理学家赫数据科学家组成的小组开发了这一系统。我们的成果远超简单分辨正面和负面情绪——这一系统甚至可以分辨出讽刺和挖苦。”

Amareos在去年推出,现在仍处于发展初期。这一服务可以分析网络对货币、股票、商品以及国家本身的情绪。而且这一系统并非只能应用于金融领域。

类似Amareos的这类工具可以快速在金融领域之外获得关注,所以其数据可以被用于很多领域,从评估一次广告宣传的效果到调查总统候选人的公众支持率。

在Amareos上申请试用账户是免费的。这样说来,对那些渴求数据的人而言,更是有必要试一试这个服务。我注册了一个账户,并在当时看到,世界对咖啡豆的情感有一点负面,但元素铀则很受欢迎。同时,人们对天然气价格很伤心,但却很喜爱棕榈油。

我尝试了很多事物,从商品到国家,然后就看到了这一幅图表:

 2 第2张

阿根廷今天看起来不错

以今天为例,世界为爱沙尼亚的前景而兴奋——只要他们对加纳和巴西有同样的看法。

这令人上瘾。一定要小心,只要你注册了Amareos的服务,就会不知不觉地花上好几个小时观察世界对保加利亚和阿根廷比索是怎么看的。

这仅仅是个开始

Amareos的服务既是B2C的,也是B2B的。一方面,它提供单用户授权,使得交易者可以利用该服务提供的海量信息。另一方面,Amareos与UBS等大型机构客户合作,向他们提供特定的大数据以及人工智能解决方案。

Philippe表示未来Amareos团队计划将8000支股票纳入分析范围,并拓展该服务支持的语言。

眼下,Amareos仅仅分析英文的新闻与社交媒体信息——其将很快推出中文与日文用户界面,但仍只能提供英文来源的信息。展望未来,Philippe希望让人工智能能够理解中文及其他语言,以此更好地理解这个世界的情感。

当然,说来容易做来难。增加中文支持不只是把软件翻译成另一种语言那么简单——这需要同样深刻的知识,需要为新的语言再次开发一套能够理解讽刺与挖苦的人工智能系统。

但Philippe相信Amareos能够成功,其他人似乎也同样乐观。这家公司已经成为香港创业圈的关注点,并获得了进入百度与渣打支持的孵化器Fintech Supercharger以及UBS支持的加速器Future of Fintech的机会。

Amareos在成为家喻户晓的公司之前还有很长的路要走,但很明显这家公司经开始成为一家重要的公司。

“让一个人阅读美国一天中发布的所有文章需要两年”,Philippe说,“但我们可以用机器在几秒钟内完成。”

Sometime in the mid-2000s, hedge fund owners got an idea. Financial markets are impacted by the news – so if you can thoroughly understand the news, and do it quickly, you can have a headstart on changes in the market. Using artificial intelligence, they devised software that could instantly analyze breaking news, assess its meaning and, within seconds, make trades based on that analysis.

Like many things in the world of hedge funds, this AI remained a toy for billionaires and professional investors. But Amareos, a Hong Kong startup, is looking to bring it to the people.

Taking the world’s temperature

“We are trying to democratize big data,” says Philippe El-Asmar, co-founder of Amareos.

Amareos is about as “big data” as you can get. The service analyzes more than one million websites and a plethora of social media networks. Using an artificial intelligence developed by partners Thomson Reuters and Market Psych, Amareos takes the temperature of the internet on specific issues.

Interested in global sentiment on the US dollar? Amareos can analyze the global chatter on the currency and present users with a detailed breakdown – how many people are angry about it, how many are trustful, how many are disgusted.

1 第1张
Today’s graph of how the internet feels about the US dollar.
“I can tell you how much joy there is on the S&P 500 and how people feel about the renminbi,” Philippe says. “A group of psychologists and data scientists worked to develop this. We go further than positive or negative – it can even detect irony and sarcasm.”

Amareos was launched last year, and is still in its early days. The service can analyze feelings on currencies, equities, commodities, and countries themselves. But there is no reason that it needs to stay in the world of finance.

“Right now we’re only focusing on financial markets, but what we’re doing could be applied to anything – politics, PR, marketing,” says Philippe.

It’s easy to see how a tool like Amareos could quickly outgrow its financial sector focus – its data would be useful for anything from testing the impact of an ad campaign to polling the public perceptions of a presidential candidate.

It is free to open up a trial account on Amareos, and for the data curious, the service is a must-try. I signed up for an account and was, in moments, seeing that the global sentiment for coffee beans is a bit negative, but uranium is wildly popular. People are sad about the prices of natural gas, but they love palm oil.

I switch from commodities to countries, and I’m greeted by this map:

2 第2张

Argentina is lookin’ good today.

As of today, the world is really excited about Estonia’s prospects – if only the same could be said for Ghana and Brazil.

It’s addicting. Beware that if you sign up for Amareos, you may suddenly spend hours seeing how the world feels about Bulgaria or the Argentine peso.

Just getting started

Amareos works as both a business-to-consumer (B2C) and business-to-business (B2B) operation. On one hand, it has single-user licenses that enable traders to tap into the service’s wealth of information. On the other, Amareos works with large institutional clients like UBS to provide specific big data and AI solutions.

Moving forward, Philippe says the Amareos team plans to index some 8,000 stocks for analysis, and also expand the company’s language capabilities.

For now, Amareos only analyzes news and social media chatter in English – it is soon rolling out a Mandarin and Japanese user interface, but those still rely on data pulled from English-language sources. Looking down the road, Philippe wants to expand the AI’s ability to understand Chinese and other languages, to provide an even more thorough understanding of the world’s sentiments.

Of course, much of this is easier said than done. Adding Chinese language support isn’t as easy as translating an app from one language to another – it requires the same kind of deeply knowledgeable, irony-and-sarcasm-understanding AI to be developed for a completely different language.

But Philippe is confident that Amareos can pull it off, and others seem optimistic as well. The company has become a hot topic in the Hong Kong startup scene, gaining admission to the Baidu- and Standard Chartered-backed Fintech Supercharger as well as the UBS-backed Future of Fintech Accelerator.

Amareos still has some time to go before it’s a household name, but it’s clear to see that the company has tapped into something big.

“It would take two years to read all of the articles posted in one day in the US,” says Philippe. “We can have the machine do it in a few seconds.”


用微信扫描可以分享至好友和朋友圈

评论

  • 钢笔水在 2016/03/05 21:27回复

    这也是不错的idea。

  • 时间的掌控在 2016/03/04 12:55回复

    ​人生就像马拉松,获胜的关键不在于瞬间的爆发,而在于途中的坚持。 非诚勿视,非己勿删。 任何一个散户炒股最开心的事情就是能够买到强势股,享受个股直线拉升的快感,都想在短期内取得非常可观的收益 。◇◆ 【二叁舞久舞七久衣巴零】

发表评论

发表评论

您的评论提交后会进行审核,审核通过的留言会展示在下方留言区域,请耐心等待。

评论

您的个人信息不会被公开,请放心填写! 标记为的是必填项

取消

薛凡未央青年

38
总文章数

TA还没写个人介绍。。。

Fintech企业申请ILC受质疑,只因传统机构害怕竞争

潘, 妍媛 2小时前

如何结合大数据科技做好汽车金融风控

车贷管家 | 杭州隼目信... 1天前

孙国峰:金融大数据应用的风险与监管

清华金融评论 1天前

第二届中国金融科技大会主题论坛:金融大数据的发展和挑战

小未 1天前

金融科技对银行的改造之路能走多远?

杨荣 | eResearch杨... 1天前

版权所有 © 清华大学五道口金融学院互联网金融实验室 | 京ICP备17044750号-1