likicode/DataWareHouse

Data Warehouse Course Project

The final project for Data Warehouse in 2015

Projects Requirements

• Data Resource
  • http://snap.stanford.edu/data/web-Movies.html
• Data needs to be stored
  • movie ID
  • comment user's ID
  • comment user's Profilename
  • comment user's Helpfulness
  • comment score for each user
  • comment time
  • comment summary
  • comment text
  • movie actors
  • movie show time
  • movie genre
  • movie director
  • movie starring actors
  • movie version
• Most frequent Research
  • Check for time
    • the number of ovies in XX year, xx month, xx season
    • how much new movies have been shown on Tuesday
  • check by movie name
    • how much versions a movie may possess
  • check by directors/ actors/ genres
  • combines research

The project implementation Process

Step 1.

• Process Data on Amazon
  • write scrawl script with python and simple bash
    • get 230 thousand items on Amazon with three servers running multiple threads at one night.
  • clean data
    • done with help of http://www.crummy.com/software/BeautifulSoup/
    • get the data need to be stored from raw html

Step 2.

• design storing plan

  • the logical storing plan: ERD
  • the physical storing plan : database table design

• Build Hive clusters

  • I bought three servers temporarily from https://www.digitalocean.com/ playing the role of namenode, edgenode and datanode. Their configuration informations are as follows:

    

    

    

  • Edgenode can keep watch on the performance of the cluster.

  • We need to compare the time consumed by Mysql and HDFS for one complex search.

Step 3.

• Research by multiple conditions and add condition automatically

  

• For example, search for the thrillers in season 1,2015

  

• The result showing in table

  

• You can click on any item to have a further search

  

• displayed time consumed comparation between mysql and HDFS in histogram and pie charts.

  

Development Tools

• platform : Mac OSX
• ETL Tool : http://www.pentaho.com/
• Hive : three servers from https://www.digitalocean.com/
  Thanks for https://www.digitalocean.com/community/tutorials/how-to-install-hadoop-on-ubuntu-13-10 and stackoverflow tips!
• Mysql : configure one on one of the server

Conclusions

• skills about scrawling loads of data online
• ETL skills
• build Hive clusters
• display research result, maybe with https://www.joomla.org/