CSC/ECE 517 Fall 2013/ch1 1w46 ka

Data Mining in Rails Application

Data mining (the analysis step of the "Knowledge Discovery in Databases" process, or KDD)is the computational process of discovering patterns in large data sets involving methods at the intersection of artificial intelligence, machine learning, statistics, and database systems. The overall goal of the data mining process is to extract information from a data set and transform it into an understandable structure for further use.

Ruby on Rails, often simply Rails, is an open source web application framework which runs on the Ruby programming language Data mining techniques like k-means clustering can be developed on ruby on rails by using various gems which ruby provides. Thus it becomes easier for a data mining analyst to write mining code in a ruby on rails application.

Weka and Ruby

Weka is a collection of machine learning algorithms for data mining tasks. The algorithms can either be applied directly to a dataset or called from an embedded code. Weka contains tools for data pre-processing, classification, regression, clustering, association rules, and visualization. It is also well-suited for developing new machine learning schemes. Weka is written in Java however it is possible to use Weka’s libraries inside Ruby. To do this, we must install the Java, Rjb, and of course obtain the Weka source code. We use JRuby and this is illustrated as follows:

Clustering Data using WEKA from jRuby

jRuby provides easy access to Java classes and methods, and WEKA is no exception. The following program builds a simple kmeans clusterer on a supplied input file, and then prints out the assigned cluster for each data instance. The 'include_class' statements are there to simplify references to classes in the API. When classifying each instance, we must watch for the exception thrown in case a classification cannot be made. Finally, notice that the filename is passed as a command-line parameter: the parameters after the name of the jRuby program are packaged up into ARGV in the usual ruby style. Assuming weka.jar, jruby.jar, and your program are in the same folder, a sample Ruby example is shown bellow:

 # Weka scripting from jruby
 require "java"
 require "weka"
 include_class "java.io.FileReader"
 include_class "weka.clusterers.SimpleKMeans"
 include_class "weka.core.Instances"
 
 # load data file
 file = FileReader.new ARGV[0]
 data = Instances.new file
 
 # create the model
 kmeans = SimpleKMeans.new
 kmeans.buildClusterer data
 
 # print out the built model
 print kmeans
 
 # Display the cluster for each instance
 data.numInstances.times do |i|
 cluster = "UNKNOWN"
  begin
   cluster = kmeans.clusterInstance(data.instance(i))
   rescue java.lang.Exception
  end
 puts "#{data.instance(i)},#{cluster}"
 end

We can see that the WEKA api makes it easy to pass in a data file. Data can be in a number of formats, including ARFF and CSV. When run on the weather.arff example (in WEKA's 'data' folder), the output looks like the following:

 Number of iterations: 3
 Within cluster sum of squared errors: 16.237456311387238
 Missing values globally replaced with mean/mode
 
 Cluster centroids:
                            Cluster#
 Attribute      Full Data          0          1
                     (14)        (9)        (5)
 
 outlook            sunny      sunny   overcast
 temperature      73.5714    75.8889       69.4
 humidity         81.6429    84.1111       77.2
 windy              FALSE      FALSE       TRUE
 play                 yes        yes        yes

Advantages of using Weka from jRuby

One of the advantages of using a language like jruby to talk to WEKA is that we should have more control on how our data is constructed and passed to the machine-learning algorithms. A good start is how to construct our own set of instances, rather than reading them directly in from file. There are some quirks to WEKA's construction of a set of Instances. In particular, each attribute must be defined through an instance of the Attribute class. This class gives a string name to the attribute and, if the attribute is a nominal attribute, the class also holds a vector of the nominal values. Each instance can then be constructed and added to the growing set of Instances. The code below shows how to 'by-hand' construct a dataset which can then be passed to one of WEKA's learning algorithms.

Example: Mining Twitter data

The following sections present a few scripts for collecting and presenting data available through the Twitter API. These scripts focus on simplicity, but you can extend and combine them to create new capabilities.

Twitter User Information

A large amount of information is available about each Twitter user. This information is only accessible if the user isn't protected. Let's look at how you can extract a user's data and present it in a more convenient way. Listing below presents a simple Ruby script to retrieve a user's information (based on his or her screen name), and then emit some of the more useful elements. You use the to_s Ruby method to convert the value to a string as needed. Note that you first ensure that the user isn't protected; otherwise, this data wouldn't be accessible.

#!/usr/bin/env ruby
require "rubygems"
require "twitter"

screen_name = String.new ARGV[0]

a_user = Twitter.user(screen_name)

if a_user.protected != true

 puts "Username   : " + a_user.screen_name.to_s
 puts "Name       : " + a_user.name
 puts "Id         : " + a_user.id_str
 puts "Location   : " + a_user.location
 puts "User since : " + a_user.created_at.to_s
 puts "Bio        : " + a_user.description.to_s
 puts "Followers  : " + a_user.followers_count.to_s
 puts "Friends    : " + a_user.friends_count.to_s
 puts "Listed Cnt : " + a_user.listed_count.to_s
 puts "Tweet Cnt  : " + a_user.statuses_count.to_s
 puts "Geocoded   : " + a_user.geo_enabled.to_s
 puts "Language   : " + a_user.lang
 if (a_user.url != nil)
   puts "URL        : " + a_user.url.to_s
 end
 if (a_user.time_zone != nil)
   puts "Time Zone  : " + a_user.time_zone
 end
 puts "Verified   : " + a_user.verified.to_s
 puts
 
 tweet = Twitter.user_timeline(screen_name).first
 
 puts "Tweet time : " + tweet.created_at
 puts "Tweet ID   : " + tweet.id.to_s
 puts "Tweet text : " + tweet.text
 
end

Twitter User Behavior

Twitter contains a large amount of data that you can mine to understand some elements of user behavior. Two simple examples are to analyze when a Twitter user tweets and from what application the user tweets. You can use the following two simple scripts to extract and visualize this information. Listing below presents a script that iterates the tweets from a particular user (using the user_timeline method), and then for each tweet, extracts the particular day on which the tweet originated. You use a simple hash again to accumulate your weekday counts, then generate a bar chart using Google Charts in a similar fashion to the previous time zone example. Note also the use of default for the hash, which specifies the value to return for undefined hashes.

#!/usr/bin/env ruby
require "rubygems"
require "twitter"
require "google_chart"

screen_name = String.new ARGV[0]

dayhash = Hash.new

# Initialize to avoid a nil error with GoogleCharts (undefined is zero)
dayhash.default = 0

timeline = Twitter.user_timeline(screen_name, :count => 200 )
timeline.each do |t|

 tweetday = t.created_at.to_s[0..2]

 if dayhash.has_key?(tweetday)
   dayhash[tweetday] = dayhash[tweetday] + 1
 else
   dayhash[tweetday] = 1
 end

end
 
GoogleChart::BarChart.new('300x200', screen_name, :vertical, false) do |bc|
 bc.data "Sunday", [dayhash["Sun"]], '00000f'
 bc.data "Monday", [dayhash["Mon"]], '0000ff'
 bc.data "Tuesday", [dayhash["Tue"]], '00ff00'
 bc.data "Wednesday", [dayhash["Wed"]], '00ffff'
 bc.data "Thursday", [dayhash["Thu"]], 'ff0000'
 bc.data "Friday", [dayhash["Fri"]], 'ff00ff'
 bc.data "Saturday", [dayhash["Sat"]], 'ffff00'
 puts bc.to_url
end

Running the above mining script provides the result of the execution of the tweet-days script in the above Listing for the developerWorks account. As shown, Wednesday tends to be the most active tweet day, with Saturday and Sunday the least active