非结构数据的分析

[xichen...@gmail.com]

IBM收购了Vivisimo，是做非结构化数据分析的，应该主要是文本数据。Oracle去年10月收购了Endeca，也主要是做非结构化数据分析的。TED上有一组视频叫Visualizing Data，其中一个视频讲到通过照片的annotation和tag把网上关于某一旅游景点的照片拼接在一起得到关于该旅游景点最全面的视图，主要技术点在于海量图片的处理能力。还有一个视频是关于视频分析的，演讲者全程录制了自己的孩子在学习说话的成长阶段的所有活动和与人的交互，然后通过分析视频解释为什么孩子学会的是这个词而不是那个词。同样的技术可以应用到分析奥巴马的讲话中的关键词是如何传导到各个电视节目中的。

关于Vivisimo的

By combining Vivisimo’s capabilities with our big data platform, IBM will be able to help clients tackle this challenge with a highly-scalable architecture designed to marry volumes of structured and unstructured data drawn from disparate sources, including Hadoop, internal warehouses and external repositories. Additionally, by providing a single point of access to disparate data, clients will be able to extract and analyze relevant information without having to move it. This kind of quick data access, navigation and ad hoc analysis helped Airbus, a Vivisimo customer, organize two petabytes of information in complex SAP and Siebel systems. That’s what we mean by getting more value out of big data.

Thanks & Best Regards