知识模式识别

  知识模式识别

 

　　在信息大爆炸的今天，对各种信息进行分类和鉴别，乃至更高层次的自动化应用，已成为了一个热门技术，要对信息进行自动化应用，最核心的技术应当是知识模式识别，如现在流行的“精准广告”，就是知识模式识别的一种，对论文进行自动化分类也是知识模式识别的一种，通过知识模式识别得到有价值的知识模式，也可以让我们做出更好的决策，知识模式识别是属于 AI 研究领域，它将跟生物技术一样引领一场新的工业革命。从以下的观点我们可以看出知识模式识别的重要性（可能翻译得不太好）：

 

“ To understand is to perceive patterns.” - Isaiah Berlin

理解就是理解模式。

 

“ Given the fast-changing and ever increasing complex nature of the world, gaining insight into how patterns are forming and structures are developing represents the most powerful way of managing in the new economy” Winslow Farrell, How Hits Happen.

鉴于自然世界的快速变化和不断的复杂化，洞察模式是怎么形成及结构是如何建立起来是操纵新经济最强有力的途径。

 

“ Recent research on cognition shows that our minds rarely make strictly logical deductions. Instead we rely on patterns and on feelings associated with those patterns”W. Brian Arthur (A leading Economist from The Santa Fe Institute).

最近在对人类认知上的研究表明 , 我们的思维很少有进行严格的逻辑推理，相反地，我们更多的是依赖模式和跟某些模式有关的感觉。

 

“ Anyone can learn to juggle. It’s about breaking down complex patterns and manoeuvres into simple tasks. Juggling is a system of tosses and throws, of different patterns, that once broken down, understood and mastered can be put together to create something magical.” Micheal Moschen (one of the world’s greatest jugglers).

任何人都可以学会抛接戏法。方法是把复杂的模式（戏法）分解变成简单的任务（动作）。抛接戏法是一个抛和接的系统，也是一个具有不同模式的系统，一旦分解了这个模式，理解和掌握了就会创建出不可思议的戏法。

 

“ The most important lesson I’ve ever learned is to understand and to trust abstractions. If you can learn both to see and to believe in life’s underlying patterns, you can make highly informed decisions every day. For example, everyone in high tech is familiar with Moore ’s Law, which states that computer-processing power will double every 18 months. Now, Moore ’s Law isn’t a law in any physical sense, but it has driven and will continue to drive our industry’s development. Yet very few people and very few companies really take this law to heart ? because really embracing it leads to seemingly nonsensical projections. Five years ago, when I told people that we’d have the processing power that we have today, lots of them ? even those who said they believed in Moore ’s Law ? thought I was being ridiculous.” Nathan Myhrvold, former Chief Technology Officer Microsoft Corp.

我学会的最重要的教训是理解并相信那些抽象出来的理论。如果你能够学会明白并相信生活中根本的模式，你则每天能够做出非常精明的决定。例如，从事高科技的人都熟悉摩尔定律：计算机的处理速度会每 18 个月翻一倍。现在，虽然摩尔定律不是任何物理意义上的定律，但它已经并将继续引导我们工业的发展。然而很少人和很少的公司有真正地把它放在心上，因为按照这个定律会导致产生表面上无意义的规划。五年前，当我告诉人们我们（五年后）会有今天的处理速度，没有人相信，甚至那些相信摩尔定律的人都认为我是荒唐的。

 

“ The third revolution is rooted in biology and self-organizing systems the search for a 　 sense of pattern ” James Bailey.

第三次（工业）革命是源于生物技术和能够寻找模式意义的自组织系统。

 

 

“ What is important is the pattern of relationships and interactions that exists and how 　 they contribute to the system as an integrated whole ” Brian Goodwin, a British Biologist.

重要的是所出现的关联和交互的模式以及它们在这个作为集成整体的系统中的贡献。

 

 

“ In any stream of ideas, some kind of pattern will be evident. The trick is to look for

patterns.” Gary Hamel, Management Strategist.

在任何一个思想流中，有些种类的模式是显而易见的，诀窍是寻找这样的模式。

 

（以上观点来自： http://www.kikm.org/pattern_recog.htm ）