订阅

多平台阅读

微信订阅

杂志

申请纸刊赠阅

订阅每日电邮

移动应用

技术 - 社交媒体

新闻筛选:社交媒体杀手级应用终浮出水面

JP Mangalindan 2011年04月28日

将数据输入到社交图中的做法已有多年。个性化新闻筛选作为最新的新闻输出应用,它的问世尽管令人欣喜,但未免有些姗姗来迟。

    即使最漫不经心的社交网络用户也会承认,Facebook或者Twitter的应用体验,简直如一泄千里的洪流,势不可挡:亲朋好友的状态更新流和共享内容流以无情的速度大量涌入,那阵势可绝非涓涓细流,却一如奔腾不息的洪流,大有泛滥之势,无论是男人、女人还是Web爬虫,都难以招架。

    当然,人们形成这样的认识,也自有其原因:Facebook用户每个月要共享300多亿条信息,Twitter用户每周也要输出10亿条推。虽然如此海量的数字信息源源不断地产生,汇聚成庞大的信息流,但鲜有卓越的解决方案能对之进行过滤、理解以及消费。

    最近,形势开始有所转变。早在1995年,尼古拉斯•尼葛洛庞帝就预测到互联网的现状。他提出了“个性化媒体”(Daily Me,又称每日的我)的概念,该服务能提供定制化的新闻体验。但是,直到过去的1年半时间里,体现这一理念的主流产品和服务才真正开始出现。由于社交网络迅速成为人们日常生活中必不可少的组成部分,同时人们开始使用平板电脑处理越来越多的内容,一系列社交媒体开始体现出个性化媒体的思想,它们力求不断优化内容筛选服务,并因此展开了激烈的角逐。这些公司包括:Flipboard、Zite以及News.me。这或许是迄今为止业界对尼葛洛庞帝的最为直接的表达敬意吧。

    这些服务的工作原理有所不同。这一点在《财富》(Fortune)于去年举办的头脑风暴技术大会(Brainstorm Tech conference)上得到了集中体现:Flipboard收集社交网络和互联网内容,再以类似杂志的形式呈现给用户;Zite则根据用户的使用行为/习惯提供个性化的用户体验。News.me不仅借助适应性技术,提供令人赏心悦目的设计;而且,与前两者不同的是,它还采取按周或月付费订阅的形式。不管怎么说,这三者都以这样或那样的方式,将用户的各种社交网络图集成到一起。而且,每家公司对于自己的思路都有一套合理的解释。

    “我们面临的一个最大挑战在于:如何简化发现、消费、消化内容的过程,使之更易于操作。” Flipboard首席执行官迈克•麦克库伊表示。“我认为,在实现尼古拉斯•尼葛洛庞帝的原始设想方面,社交网络称得上迄今最伟大的探索,它们是用户的个性化探索和发现工具。”

    这点同样也是专注于Twitter的初创公司Sulia的核心理念。该公司由乔纳森•格里克创建,是得到Twitter支持的屈指可数的第三方服务之一。其任务是:对推和Twitter列表进行过滤,并能针对任一话题,确定哪些人加入的Twitter列表最多并且发推最多。这些发推者则依序成为五花八门的各类频道的“顶级专家”来源。这些频道包括:Sulia的各类频道,从政治与理想(Politics & Causes)到艺术与娱乐(Arts & Entertainment)等各类主题无所不包;以及其他可通过Sulia或者Flipboard等应用程序正常收看的频道。

    无论麦克库伊还是格里克都坦承,尽管Sulia等服务有助于对海量信息和内容进行过滤,但它们也不过是万里长征刚刚走完了第一步而已。面对这一艰巨挑战,根本不存在清晰见底的答案,或者至少目前他们还无法提供这样的答案。但是,他们的目标再显而易见不过:增强技术的智能化。

    格里克曾在《纽约时报》电子媒体公司(The New York Times Electronic Media Company)担任产品开发与技术部门的负责人。他表示:“我们不妨将这些信息碎片看作组成棉被的棉絮,我们所做的就是将这些碎片缝接成一张密密实实的‘大棉被’,我们可以干得更出色。”

    Zite已经开始进行这方面的尝试了。他们根据点击量、用户在某篇新闻报道上逗留的时间长短以及语义(比如“左翼”还是“右翼”网络日志作者)等若干因素,不断将相关内容输送给读者。从当前的技术水平看,这些服务尽管前景不可估量,但也并非完美无缺。比如,它们可以依据与读者在Facebook或者Twitter上进行互动的人群(同事、导师、密友以及家人)等深层视角,对内容进行筛选和组织,并且做到这一切均在后台实现,读者无需像传统做法一样,得对新闻报道进行投票表决才行。

    相反,麦克库伊认为,出色的内容筛选服务,应该无需用户输入过多的信息。无论应用的系统是计算机、平板电脑还是智能手机,总会有一部分用户希望自己有权使用所有按钮,尽可能地发掘和应用产品与服务的各种功能。但是,大多数或者主流用户基本上还是希望将所有那些操作都放在后台实现。

    就这点而言,开发人员和内容筛选人员应该把握好尺度,既能合理地满足用户需求,又不能聪明过头。如果做得过度,便会物极必反,经过一道道精挑细选的内容,实际上起到的恰是反面作用。如果内容筛选服务对我们的社交图挖掘过深,它们最终提供的信息很可能范围既过于狭窄,内容也过于类同。那样的话,我们每天打开报纸浏览新闻,突然发现奇闻趣事时的那份乐趣和惊喜,也就荡然无存了。此点至关重要,管窥一下StumbleUpon每月提供的10亿个“意外发现”,便可见一斑。但这也并非说,为读者提供意外发现的阅读乐趣,只是传统媒体的专利。如果内容的个性化程度过高,比如说,用户只能收阅有左翼或者右翼倾向的内容,结果只能是进一步强化其原有世界观,那也并非什么好事。如果用户根本无从知道相反观点,也就不存在辩论的空间和可能了。

    换言之,虽然社交筛选新闻应用程序可能极为有用,但是就其效用而言,也应该有个分界线,如果服务得过了头,新闻聚合服务所扮演的就不再是智能过滤器,而仅仅是一面镜子了。

    译者:大海

    Even the most casual social network user will admit that the Facebook or Twitter experience can be overwhelming -- that merciless stream of status updates and shared content, which sometimes feels less like a stream and more like a deluge, waits for no man, woman, or Web crawler.

    Of course, there's good reason to feel that way: Facebookers share 30-billion plus pieces of information each month, and Twitter users output 1 billion tweets weekly. There's a tremendous amount of digital information floating around and few great solutions for filtering it, making sense of it, and consuming it.

    That's changing. Nicholas Negroponte foreshadowed the current state of things back in 1995 with the "Daily Me," a customized news experience, but it's only been over the last 18 months that his idea has manifested itself via mainstream products and services. As social networks quickly become entrenched in our everyday lives and content becomes increasingly consumed on tablets, we're seeing the Daily Me embodied among competitors in a race towards better content curation: Flipboard, Zite, and News.me, maybe the most obvious homage to Negroponte yet.

    They all work differently. Unveiled at Fortune's Brainstorm Tech conference last year, Flipboard pulls social network and web content and presents it in magazine-like form; Zite tries to personalize the user experience based on behavior. And News.me meshes attractive design with adaptive technology, and differentiates itself with a paid subscription weekly or monthly model. All of them integrate your social network graphs in some way, and there's a reason for that.

    "One of the biggest challenges is how do you make that content more easily discoverable, easily consumable, easily digestible," says Flipboard CEO Mike McCue. "I think social networks are the biggest heuristic we've ever had to achieving the original vision of Nicholas Negroponte's idea -- they act as your personalization heuristics."

    That same concept is at the core of the Twitter-focused start-up Sulia. Founded by Jonathan Glick, it's one of the few Twitter-endorsed third-party services. Its mission: filter tweets and Twitter lists to figure out who the most frequently listed and prolific tweeters are on any given topic. Those tweeters in turn become the sources of "Top experts" of different broad channels on Sulia, from Politics & Causes to Arts & Entertainment, and those channels are viewable on Sulia.com proper or via apps like Flipboard.

    Though services like Sulia help filter through all the riff-raff, there's a long way to go, something McCue and Glick readily admit. It's a challenge with no clear discernible answer, or at least one they're ready to reveal yet, other than the obvious: make the technology smarter.

    "Think of these pieces of information like individual pieces of a quilt," says Glick, who once served as Head of Product Development and Technology for The New York Times Electronic Media Company. "We can do a better job of patching those pieces together into one cohesive, larger work."

    Zite already tries to do that, based on factors like click-throughs, how long users linger on stories, and even semantics -- "left-wing" blogger? "Right wing?" -- to increasingly "surface" relevant content to readers. The current level of technology remains promising but flawed. Content could be curated depending on deeper aspects like who you interact with the most on Facebook or Twitter -- colleagues, mentors, close friends and family -- and all without readers having to perform banal tasks like voting stories up or down.

    Instead, McCue argues that better content curation should come without what he views as superfluous user input. Whether it's a computer, a tablet or a smartphone, there will always be a user contingent that wants access to all the knobs and buttons to tweak the product and service as much as they like. Yet, the majority or mainstream will arguably want all that stuff tucked away behind-the-scenes.

    Developers and curators will and should get smart on this front, but not too smart. Cross the line, and heavily curated content actually becomes negative. As services dig deeper into our social graph, they run the risk of eventually presenting a narrow, homogeneous mix of information that lacks the serendipity that comes from discovering the news for ourselves, as we might when we open the day's newspaper. That's an important element -- not to say it's only the province of old-media to be serendipitous -- if StumbleUpon's 1 billion monthly "stumbles" are any indication. If content becomes so personalized that, say, users only see left-wing- or right-wing-leaning content that reinforces their world-views, that's no good either. There's no room for debate if users simply don't know what the other side is saying.

    In other words, there may be a lot of utility in socially curated news apps, but there could also be a point where rather than act as a very smart filter, news aggregators become nothing more than a simple, dumb mirror.

我来点评

相关稿件

  最新文章

最新文章:

500强情报中心

财富专栏