crawler with data analysis (Hadoop, MapReduce, HBase) - Phase I - Data Modeling

http://www.donanza.com/jobs/p3315101-crawler_with_data_analysis_hadoop_mapreduce_hbase_phase_i

crawler with data analysis (Hadoop, MapReduce, HBase) - Phase I - Data Modeling

Goal for Phase 1: given a topic in English (e.g. "skiing"), crawl the web (sites, blogs, social media) and collect 1 million relevant articles/pages/posts/documents. Perform analysis and generate meaningful reports on the topic, potentially including top keywords, concepts, related topics or concepts, Optional task (bonus): add "intelligence" to your analysis, by determining rank/reputation, sentiment (negative vs. positive), type (opinion article vs. advertisement vs. for sale ad vs. wanted ad) - we are flexible and open to ideas. Development/staging environment: 3-node cluster, CentOS 5.6 and Cloudera CDH3 (Hadoop, MapReduce, Hue, Pig, Flume, HBase) + one management machine with CDH. If you bid on this job, please describe your prior experience with Big Data, and tell us how you would approach this problem, a high-level overview of steps you will need to perform... It's important for us to see the way you approach problems. We speak English and Russian fluently. Depending on your approach, we will define milestones and timeline together. This is Phase I of the project, do your best! Desired Skills: Data Modeling, Scripts & Utilities, CentOS, Hadoop, MapReduce