Apache Mahout: 基于MapReduce框架的实时决策系统

daizhe

Apache Mahout: Real-Time Decisioning in the MapReduce Framework<p>Here is a bit of good news for the complex event processing space.&nbsp; Folks on the <a href="http://www.nabble.com/how-can-I-contribute-td21449196.html" target="_blank" target="_blank">Apache Mahout developers mailing list</a> are showing an accelerated interest in topics related to real-time decision-making, starting with a <a href="http://en.wikipedia.org/wiki/Markov_decision_process" target="_blank" target="_blank">Markov decision process</a>. The <a href="http://en.wikipedia.org/wiki/Hidden_Markov_model" target="_blank" target="_blank">Hidden Markov Model (HMM)</a> is listed under “non map-reduce algorithms” on the <a href="http://cwiki.apache.org/MAHOUT/" target="_self" target="_blank">Mahout wiki</a>.&nbsp; The developer’s interest appears to be partially motivated by this presentation, <a href="http://www.cs.cmu.edu/%7Ereids/planning/handouts/MDPs.pdf" target="_blank" target="_blank">Towards Learning in Probabilistic Action Selection: Markov Systems and Markov Decision Processes</a>, the October 29 entry in these <a href="http://www.cs.cmu.edu/%7Ereids/planning/notes.html" target="_blank" target="_blank">CMU AI Planning, Execution, and Learning Lecture Notes</a>.</p>
<p>This is great news for fans of complex event processing, moving the state-of-the-art of CEP/EP closer to a <a href="http://www.thecepblog.com/2009/01/14/it-infrastructure-capability-as-a-service/" target="_blank" target="_blank">Capability as a Service (CaaS)</a> model for real-time detection and decision-making.&nbsp;&nbsp; This overall CaaS direction also aligns nicely with <a title="A Brief Introduction to Blackboard Architectures" rel="bookmark" href="http://www.thecepblog.com/2009/01/14/2008/07/20/a-brief-introduction-to-blackboard-architectures/">the blackboard architectural construct</a>
for complex, distributed event processing classes of problems.&nbsp;
Distributed blackboard architectures can be realized in cloud computing
CaaS models.</p>
<p>It is a safe bet that my favorite real-time software engineering team, headquartered in Palo Alto (<em>wink, wink</em>),
is keeping a close watch on MapReduce-related, Apache Mahout and
Hadoop.&nbsp; After a bit of a disappointing couple of years (2007 and 2008)
in the CEP/EP technology space, with very little progress toward
scaleable real-time analytics that are useful for decision-making, I am
starting to be motivated again thanks to the application of <a href="http://en.wikipedia.org/wiki/MapReduce" target="_blank" target="_blank">Google’s MapReduce</a> framework to suitable analytics.</p>