Bahir Dar Data Mining Lab-Weka Edited.pdf

Bahir Dar University
Bahir Dar Institute of Technology
Faculty of Computing
Data Mining and Warehouse Laboratory Manual
By:
Mekuriaw Melkamu
Tizalegn Tilaye
January 2019 G.C
Bahir Dar, Ethiopia

I
Contents
1. INTRODUCTION TO DATA WAREHOUSE..............................................................................................1
1.2. WHAT IS A DATA WAREHOUSE? .................................................................................................1
1.3. DATA WAREHOUSE INTEGRATION PROCESS..............................................................................1
1.4. SETTING UP A DATA WAREHOUSE..............................................................................................2
2. LAB SESSION 1: APRIORI ALGORITHM ................................................................................................3
2.1. INTRODUCTION................................................................................................................................3
2.2. LIMITATIONS OF APRIORI ALGORITHM.......................................................................................3
2.3. PHASES OF KNOWLEDGE DISCOVERY.........................................................................................3
2.4. APPLICATION OF DATA MINING..................................................................................................3
2.5. ASSOCIATION RULES ...................................................................................................................3
2.6. SUPPORT AND CONFIDENCE.......................................................................................................3
2.7. GENERATING ASSOCIATION RULES.............................................................................................4
2.8. USE OF APRIORI ALGORITHM......................................................................................................4
2.9. CREATING FREQUENT SETS.........................................................................................................4
2.10. APRIORI ALGORITHM IN PSEUDOCODE..................................................................................4
3. LAB SESSION 2: FP GROWTH ALGORITHM..........................................................................................9
3.1. INTRODUCTION................................................................................................................................9
STEP 1: FP-TREE CONSTRUCTION............................................................................................................9
4. LAB SESSION 3: K-MEANS CLUSTERING.................................................................................................13
4.1. WHAT IS CLUSTERING? ..................................................................................................................13
4.2. K-MEANS CLASTURING...................................................................................................................13
4.3. How the K-Mean Clustering algorithm works?..............................................................................14
5. LAB SESSION 4: HIERARCHICAL CLUSTERING ........................................................................................22
5.1. HIERARCHICAL CLUSTERING ..........................................................................................................22
6. LAB SESSION 5: BAYESIAN CLASSIFICATION..........................................................................................26
6.1. BAYESIAN CLASSIFICATION ............................................................................................................26
7. LAB SESSION 6: DECISION TREE ........................................................................................................37
7.1. DECISION TREE...............................................................................................................................37
8. LAB SESSION 7: SUPPORT VECTOR MACHINE.......................................................................................43
8.1. SUPPORT VECTOR MACHINES...................................................................................................43
9. LAB SESSION 8: APPLICATIONS OF CLASSIFICATION FOR WEB MINING...............................................47
9.1 WEB MINING...................................................................................................................................47
9.1.1. WEB CONTENT MINING ..........................................................................................................47
9.1.2. WEB STRUCTURE MINING.......................................................................................................47
9.1.3. WEB USAGE MINING...............................................................................................................47
10. LAB SESSION 9: CASE STUDY ON TEXT MINING ..................................................................................49

1
1. INTRODUCTION TO DATA WAREHOUSE
1.2.WHAT IS A DATA WAREHOUSE?
Data warehouse databases are designed for query and analysis, not transactions. The data that is
collected from various sources is separated into analytic and transaction workloads while
enabling extraction, reporting, data mining and a number of different capabilities that transform
the information into actionable, useful applications.
The main data warehouse structures listed are the basic architecture, which is a simple set up
that allows end-users to directly access the data from numerous sources through the warehouse,
a second architecture is a warehouse with a staging area that simplifies warehouse management
and helps with cleaning and processing the data before it is loaded into the arehouse. And finally
there is the architecture with both a staging area and a data mart. Data marts are used to create
and maintain custom categories in organizations with specialized database designed for specific
businesses, so for example if an organization had a data warehouse for sales, separate from
advertising, then the data mart setup would best serve their needs. To further understand a
data warehouse, it is important to look at its characteristics, which are subject orientation,
integration, non-volatility, and time variance.
Subject Oriented: This refers to when data is giving information on a particular subject. For
example, a company could use data to analyze their company‘s marketing data, and it‘s
effectiveness. The devotion of a data warehouse to a specific matter is the key component of a
subject-oriented warehouse.
Integrated: This is when data gathered from a number of disparaging sources, and then all
gathered into a coherent whole. By organizing differing data into consistent formats, companies
can resolve problems and inconsistencies among units of measurement and promote better
results.
Nonvolatile: This refers to data that remains stable and unaffected by new developments. Once
entered into the system, this data should not change, ensuring comparison analysis over a long
period of time.
Time Variant: This refers to data gathered is identified with a particular time period and
focuses on change over time. By using a large amount of data, spread over a long time
period, analysts can decipher patterns, and business relationships that would have otherwise been
overlooked.
1.3.DATA WAREHOUSE INTEGRATION PROCESS
The whole purpose of data mining is to facilitate business analysis. And to accomplish this, raw
data must be arranged and consolidated into an information base usable by the firm. This process
is referred to as ETL (Extract, Transform, & Load), which though it may seem like specified
steps, is in opposition referring to a broader concept.
EXTRACTION
This step in the process refers to removing the data from its source and making it
accessible for further processing. All the needed data is retrieved without affecting the source
system‘s performance, response time or locking in a negative manner. This first step in the ETL
process usually involves a cleaning phase in which data quality is ensured through data
unification. The rules of unification should include things such as making identifiers unique such
as gender categories, phone number, and zip code conversions into standard form and validation
of address fields converted into the proper format.

2
TRANSFORMATION
This step applies a set of rules to change source data into similar dimensions so the same units of
measurement can be used. This transformation step also joins data from a variety of sources,
generates aggregates, surrogate keys and applies validation and new values.
LOADING
The loading phase is a two-part process of disabling constraints and indexes before the load
process starts and then enables them once the load is completed. In this step, the target of the load
process is often a database.
1.4.SETTING UP A DATA WAREHOUSE
The main purpose of a data warehouse is to organize large amounts of stable data to be easily
retrieved and analyzed. So when setting up, care must be taken to ensure the data is rapidly
accessible and easily analyzed. One way of designing this system is with the use of dimensional
modeling, which allows large volumes of data to be efficiently queried and examined. Since much
of the data in warehouses is stable, that is, unchanging, there is hardly a need for repetitive backup
methods. Also, once new data is loaded it can be updated and backed up right away by way of,
in some cases, the data preparation database, so it becomes available for easy access. There are
four categories of data warehousing tools; these are extraction, table management, query
management and data integrity tools. All these tools can be used in the setup and maintenance of
the best technology to manage and store the huge amounts of data a company collects, analyzes
and reviews.
COMPANY ANALYSIS
The first step, in setting up the company‘s data warehouse, is to evaluate the firm’s
objectives, For example, a growing company might set the objective to engage customers in
building rapport. By examining what the company needs to do to achieve these tasks, what will
need to be tracked, the key performance indicators to be noted and a numeric evaluation of the
company‘s activities the company can note and evaluate where they need to get started.
EXISTING SYSTEM ANALYSIS
By asking customers and various stakeholders pointed questions, Business Intelligence leaders
can gather the performance information they currently have in place that is or isn‘t effective.
Reports can be collected from various departments in the company, and they may even be able
to collect analytical and summary reports from analysts and supervisors. INFORMATION
MODELING OF CORE BUSINESS PROCESSES
An information model is conceptual, and allows for one to form ideas of what business processes
need to be interrelating and how to get them linked. Since the data warehouse is a collection of
correlating structures, creating a concept of what indicators need to be linked together to create
top performance levels is a vital step in the information modeling stage. A simple way to design
this model is to gather key performance indicators into fact tables and relate them to dimensions
such as customers, salespeople, products and such.
DESIGN AND TRACK
Once all those concepts are set in place, the next critical step is to move data into the warehouse
structure and track where it comes from and what it relates to. In this phase of design, it is crucial
to plan how to link data in the separate databases so that the information can be connected as it
is loaded into the data warehouse tables. The ETL process can be pretty complex and require
specialized programs with sophisticated algorithms, so the right tools have to be chosen at the
right, and most cost effective price for the job. Because the data is to be tracked over time, the
data will need to be available for a very long period. However the grain (atoms or make up) of

3
the data will defer over time, but the system should be set that the differing granularity is still
consistent throughout the singular data structure.
2. LAB SESSION 1: APRIORI ALGORITHM
AIM:
This experiment illustrates some of the basic elements of association rule mining using WEKA.
The sample dataset used for this example is apriori.arff.
2.1. INTRODUCTION
Developed by Agrawal and Srikant 1994
Innovative way to find association rules on large scale, allowing implication outcomes
that consist of more than one item
Based on minimum support threshold
Three versions:
1. Apriori (basic version) faster in first iterations
2. AprioriTid faster in later iteratons
3. AprioriHybrid can change from Apriori to AprioriTid after first iterations
2.2.LIMITATIONS OF APRIORI ALGORITHM
 Needs several iterations of the data Uses a minimum support threshold
 Difficulties to find rarely occurring events
 Alternative methods (other than appriori) can address this by using a minimum
support thresold
 Some competing alternative approaches focus on partition and sampling.
2.3.PHASES OF KNOWLEDGE DISCOVERY
 Data selection Data cleansing
 Data enrichment (integration with additional resources) Data transformation or
encoding
 Data mining
 Reporting and display (visualization) of the discovered knowledge
2.4.APPLICATION OF DATA MINING
 Data mining can typically be used with transactional databases (for ex. in shopping cart
analysis)
 Aim can be to build association rules about the shopping events
 Based on item sets, such as {milk, cocoa powder} 2-itemset
{milk, corn flakes, bread} 3-itemset
2.5.ASSOCIATION RULES
 Items that occur often together can be associated to each other
 These together occuring items form a frequent itemset
 Conclusions based on the frequent itemsets form association rules
 For ex. {milk, cocoa powder} can bring a rule cocoa powderèmilk
2.6.SUPPORT AND CONFIDENCE
 If confidence gets a value of 100 % the rule is an exact rule
 Even if confidence reaches high values the rule is not useful unless the support value is
high as well

4
 Rules that have both high confidence and support are called strong rules
 Some competing alternative approaches can generate useful rules even with low
support values
2.7.GENERATING ASSOCIATION RULES
• Usually consists of two sub problems:
1) Finding frequent itemsets whose occurrences exceed a predefined minimum
support threshold
2) Deriving association rules from those frequent itemsets (with the constrains
of minimum confidence threshold)
• These two sub problems are solved iteratively until new rules no more emerge
• The second sub problem is quite straight- forward and most of the research focus
is on the first sub problem.
2.8.USE OF APRIORI ALGORITHM
• Initial information: transactional database D and user-defined numeric minimum
support
threshold min_sup
• Algorithm uses knowledge from previous iteration phase to produce frequent itemsets
• This is reflected in the Latin origin of the name that means ‖from what comes before.
2.9.CREATING FREQUENT SETS
• Let‘s define:Ckas a candidate itemset of size kLk as a
frequent itemset of size k
• Main steps of iteration are:
1) Find frequent set Lk-1
2) Join step: Ck is generated by joining Lk-1 with itself (Cartesian product
Lk-1 x Lk-1)
3) Prune step (apriori property): Any (k − 1) size itemset that is not frequent
cannot be a subset of a frequent k size itemset, hence should be removed
4) Frequent set Lk has been achieved.
Algorithm uses breadth-first search and a hash tree structure to make candidate itemsets
efficiently
Then occurrence frequency for each candidate itemset is counted
Those candidate itemsets that have higher frequency than minimum support threshold
are qualified to be frequent itemsets.
2.10. APRIORI ALGORITHM IN PSEUDOCODE
L1= {frequent items};
for(k= 2; Lk-1!=∅; k++) do begin
Ck= candidates generated from Lk-1 (that is: Cartesian product Lk-1 x Lk-1 and
eliminating
any k-1 size itemset that is not frequent); for each transaction t in database do
increment the count of all candidates in Ck that are contained in t Lk =
candidates in Ck with min_sup
end
return kLk;

5
STEPS:
1.Open Excel and prepare dataset and save as apriori.csv
2. Open weka tool and click Explorer button.

6
Click open file..button in Preprocess tab and select apriori.csv.
Remove- Transaction field and save the file as aprioritest.arff

7
Goto Associate tab – choose Apriori and click Start button.
OUTPUT:

8
The above screenshot shows the association rules that were generated when apriori algorithm is
applied on the given dataset.

9
3.LAB SESSION 2: FP GROWTH ALGORITHM
AIM:
This experiment illustrates the use of FP-Growth associate in weka. The sample data set used in this
experiment is apriori.arff. This document assumes that appropriate data preprocessing has been
performed.
3.1. INTRODUCTION
Apriori: uses a generate-and-test approach – generates candidate itemsets and tests if they
are frequent.
– Generation of candidate itemsets is expensive (in both space and time)
– Support counting is expensive
• Subset checking (computationally expensive)
• Multiple Database scans (I/O)
FP-Growth: allows frequent itemset discovery without candidate item set generation. Two
step approach:
– Step 1: Build a compact data structure called the FP-tree.
Built using 2 passes over the data-set.
– Step 2: Extracts frequent item sets directly from the FP-tree
STEP 1: FP-TREE CONSTRUCTION
Ø
FP-Tree is constructed using 2 passes over the data-set: Pass 1: Scan
data and find support for each item.
Discard infrequent items.
Sort frequent items in decreasing order based on their support. Pass 2: Nodes
correspond to items and have a counter
1. FP-Growth reads 1 transaction at a time and maps it to a path
– Fixed order is used, so paths can overlap when transactions share items. In this case,
counters are incremented
2. Pointers are maintained between nodes containing the same item, creating singly linked
lists (dotted lines)
– The more paths that overlap, the higher the compression. FP-tree may fit in memory.
3. Frequent item sets extracted from the FP-Tree.
PROCEDURE:
1. Open the data file in Weka Explorer. It is presumed that the required data fields have been
discretized.
2. Clicking on the associate tab will bring up the interface for association rule algorithm.

10
3. We will use FP-Growth algorithm.
4. In order to change the parameters for the run (example support, confidence etc) we click on the
text box immediately to the right of the choose button.
STEPS:
1.Open Excel and prepare dataset and save as - apriori.csv
2. Open weka tool and click Explorer button.
Click open file..button in Preprocess tab and choose apirior.csv.

11
Remove- Transaction field and save the file as aprioritest.arff.
Goto Associate tab – choose FPGrowth and click Start button.

12
OUTPUT:
FP-Growth found 2 rules (displaying top 2)

13
4. LAB SESSION 3: K-MEANS CLUSTERING
AIM:
This experiment illustrates the use of simple k-mean clustering with Weka explorer. The sample
data set used for this example is based on the vote.arff data set. This document assumes that
appropriate pre-processing has been performed.
4.1. WHAT IS CLUSTERING?
• Organizing data into classes such that there is
high intra-class similarity
low inter-class similarity
• Finding the class labels and the number of classes directly from the data (in contrast to
classification).
• More informally, finding natural groupings among objects.
4.2. K-MEANS CLASTURING
K-Means is simplest unsupervised learning algorithms that solve the well-known
clustering problem. The procedure follows a simple and easy way to classify a given data set
through a certain number of clusters (assume k clusters) fixed apriori. The main idea is to
define k centers, one for each cluster. These centers should be placed in a cunning way because
of different location causes different result. So, the better choice is to place them as much as
possible far away from each other. The next step is to take each point belonging to a given data
set and associate it to the nearest center. When no point is pending, the first step is completed and
an early group age is done.
 The k-means algorithm is an algorithm to cluster n objects based on attributes into
partitions, where k<n.
 It is similar to the expectation-maximization algorithm for mixtures of Gaussians in that
they both attempt to find the centers of natural clusters in the data.
 It assumes that the object attributes form a vector space.
 Simply speaking k-means clustering is an algorithm to classify or to group the objects
based on attributes/features into K number of group.
 K is positive integer number.
 The grouping is done by minimizing the sum of squares of distances between data and the
corresponding cluster centroid.

14
4.3. How the K-Mean Clustering algorithm works?
Step 1: Begin with a decision on the value of k = number of clusters.
Step 2: Put any initial partition that classifies the data into k clusters. The training samples randomly
or systematically as the following:
1. Take the first k training sample as single-element clusters
2. Assign each of the remaining (N-k) training samples to the cluster with the nearest centroid.
After each assignment, recomputed the centroid of the gaining cluster.
Step 3: Take each sample in sequence and compute its distance from the centroid of each of the
clusters. If a sample is not currently in the cluster with the closest centroid, switch this sample to
that cluster and update the centroid of the cluster gaining the new sample and the cluster losing
the sample.
Step 4: Repeat step 3 until convergence is achieved, that is until a pass throughthe training
sample causes no new assignments.
PROCEDURE:
1. Run the Weka explorer and load the data file vote.arff in preprocessing interface.
2. In order to perform clustering select the ‘cluster’ tab in the explorer and click on the choose
button. This step results in a dropdown list of available clustering algorithms.
3. In this case we select ‘simple k-means’.
4. Next click in text button to the right of the choose button to get popup window shown in the
screenshots. In this window we enter six on the number of clusters and we leave the value of the
seed on as it is. The seed value is used in generating a random number which is used for making
the internal assignments of instances of clusters.
5. Once of the option have been specified. We run the clustering algorithm there we must make
sure that they are in the ‘cluster mode’ panel. The use of training set option is selected and then
we click ‘start’ button. This process and resulting window are shown in the following
screenshots.

15
STEPS: (Using Weka Explorer)
1.Open weka tool and click Explorer.
2. clickOpen file…in Preprocess tab.- choose vote.arff

16
Choose cluster tab – click choose button – choose SimpleKmeans.
Click Start button.

18
Goto - Visualize tab- click one box any visualize.
In the above move the jitter to last and to view the results of clustering

19
STEPS: (Using WekaKnowledgeFlow)
1.Open weka tool and click Explorer.
Click DataSources in the left side window and choose ArffLoader and draw in right side
window.
Double click ArffLoader and choose the file name.
Similarly do the following.

21
Right click the TextViwer choose Show results option.
OUTPUT:

22
5. LAB SESSION 4: HIERARCHICAL CLUSTERING
AIM:
This experiment illustrates the use of one hierarchical clustering with Weka explorer. The sample
data set used for this example is based on the vote.arff data set. This document assumes that
appropriate pre-processing has been performed.
5.1. HIERARCHICAL CLUSTERING
Hierarchical clustering is a method of cluster analysis which seeks to build a hierarchy of
clusters. Strategies for hierarchical clustering generally fall into two types. Agglomerative is a
"bottom up" approach: each observation starts in its own cluster, and pairs of clusters are merged
as one moves up the hierarchy. Divisive is a "top down" approach: all observations start in one
cluster, and splits are performed recursively as one moves down the hierarchy.
PROCEDURE:
discretized.Clicking on the cluster tab will bring up the interface for cluster algorithm.
2. We will use hierarchical clustering algorithm.
3. Visualization of the graph
STEPS:
The following screenshot shows the clustering rules that were generated when hierarchical
clustering algorithm is applied on the given dataset.
1.OpenWeka tool and choose Explorer.

23
2.Click - Open file… in preprocess tab –choose vote.arff.
3.Goto Cluster tab – click choose button - select HierarchicalClusterer.
5.Click Start button.

24
OUTPUT:
6. Visualize the tree by right clicking and choose Visualize Tree option.

26
6. LAB SESSION 5: BAYESIAN CLASSIFICATION
AIM:
This experiment illustrates the use of Bayesian classifier with Weka explorer. The sample data set
used for this example is based on the weather.nominal.arff data set. This document assumes
that appropriate pre-processing has been performed.
6.1. BAYESIAN CLASSIFICATION
Bayesian classification is based on Bayes theorem. Bayesian classifiers are the statistical
classifiers. Bayesian classifiers can predict class membership probabilities such as the probability
that a given tuple belongs to a particular class.
Bayesian Classification: Why?
A statistical classifier: performs probabilistic prediction, i.e., predicts class membership
probabilities
Foundation: Based on Bayes‘ Theorem.
Performance: A simple Bayesian classifier, naïve Bayesian classifier, has comparable
performance with decision tree and selected neural network classifiers
Incremental: Each training example can incrementally increase/decrease the probability
that a hypothesis is correct — prior knowledge can be combined with observed data.
Standard: Even when Bayesian methods are computationally intractable, they can provide
a standard of optimal decision making against which other methods can be measured
PROCEDURE:
discretized.
2. Next we select the “classify” tab and click choose button to select the “NavieBayes” in the
classifier.
3. Now we specify the various parameters. These can be specified by clicking in the text box to
the right of the chose button. In this example, we accept the default values his default version
does perform some pruning but does not perform error pruning.
4. We select the 10-fold cross validation as our evaluation approach. Since we don’t have
separate evaluation data set, this is necessary to get a reasonable idea of accuracy of generated
model.
5. We now click start to generate the model .the ASCII version of the tree as well as evaluation
statistic will appear in the right panel when the model construction is complete.
6. Note that the classification accuracy of model is about 69%.this indicates that we may find
more work. (Either in preprocessing or in selecting current parameters for the classification)
7. Now weka also lets us a view a graphical version of the classification tree.

27
8. We will use our model to classify the new instances.
Bayesian Classification - For Training and Testing
How do I divide a dataset into training and test set?
You can use the Remove Percentage filter (package weka.filters.unsupervised.instance). In the
Explorer just do the following:
Training set:
Load the full dataset
select the RemovePercentage filter in the preprocess panel set
the correct percentage for the split
apply the filter
save the generated data as a new file
Test set:
Load the full dataset (or just use undo to revert the changes to the dataset) select
the RemovePercentage filter if not yet selected
set the invertSelection property to true
apply the filter
save the generated data as new file
STEPS: (For Weka Explorer)
1.Open Weka Tool and click Explorer button.
2. Open file in preprocess tab - weather.nominal.arff

28
Click Choose button in Preprocess tab.

29
Click the above RemovePercentage –P 50.0 and change Percentage as 60.0 – click OK.

30
Click Apply and Save – Type file name weather.nominaltest.arff
Again open file - weather.nominal.arff.

31
Click Choose button in Preprocess tab.

32
To change the InverSelection as True – click OK.
Click Apply and Save – Type file name weather.nominaltraining.arff.
Apply NavieBayes classification using Training set. (weather.nominaltraining.arff)
Goto Classify tab in Weka Explorer and click Choose button- select NavieBayes.

34
Apply NavieBayes classification using Test set. (weather.nominaltest.arff)
Now click Supplied test set – Set button – click Open file.. –choose - weather.nominaltest.arff.
Click Start button.

35
OUTPUT:
STEPS: (For Knowledge Flow)
Click Knowledge Flow in Weka GUI Chooser.
To create the following and click Run button and Right click the TextViwer and select Show
Result.
Draw ArrfLoader and select the filename.

37
7.LAB SESSION 6: DECISION TREE
AIM:
This experiment illustrates the use of j-48 classifier in weka. The sample data set used in this
experiment is weather dataset available at arff format. This document assumes that appropriate data
preprocessing has been performed.
7.1. DECISION TREE
A decision tree is a structure that includes a root node, branches, and leaf nodes. Each
internal node denotes a test on an attribute, each branch denotes the outcome of a test, and each
leaf node holds a class label. The topmost node in the tree is the root node. The following
decision tree is for the concept buy computer that indicates whether a customer at a company is
likely to buy a computer or not. Each internal node represents a test on an attribute. Each leaf
node represents a class.
The benefits of having a decision tree are as follows
 It does not require any domain knowledge.
 It is easy to comprehend.
 The learning and classification steps of a decision tree are simple and fast.
Decision Tree Induction Algorithm
A machine researcher named J. Ross Quinlan in 1980 developed a decision tree algorithm
known as ID3 (Iterative Dichotomiser). Later, he presented C4.5, which was the successor of
ID3. ID3 and C4.5 adopt a greedy approach. In this algorithm, there is no backtracking; the trees
are constructed in a top-down recursive divide-and-conquer manner.
Algorithm :Generate_decision_tree
Input:
 Data partition, D, which is a set of training tuples and their associated class labels.
 attribute_list, the set of candidate attributes.
 Attribute selection method, a procedure to determine the splitting criterion that best
partitions that the data tuples into individual classes.
 This criterion includes a splitting_attribute and either a splitting point or splitting subset.
Output:
A Decision Tree

38
Method
Create a node N;
if tuples in D are all of the same class, C then return
N as leaf node labeled with class C;
if attribute_list is empty then
return N as leaf node with labeled with majority class in D;|| majority voting
applyattribute_selection_method(D, attribute_list)
to find the best splitting_criterion;
label node N with splitting_criterion;
if splitting_attribute is discrete-valued and
multiway splits allowed then // no restricted to binary trees
attribute_list = splitting attribute; // remove splitting attribute for
each outcome j of splitting criterion
// partition the tuples and grow subtrees for each partition
Let Dj be the set of data tuples in D satisfying outcome j; // a partition
If Dj is empty then
attach a leaf labeled with the majority
class in D to node N;
else
end for
attach the node returned by Generate decision
tree(Dj, attribute list) to node N;
return N;
Tree Pruning
Tree pruning is performed in order to remove anomalies in the training data due to noise or
outliers. The pruned trees are smaller and less complex.
Tree Pruning Approaches
Here is the Tree Pruning Approaches listed below –
 Pre-pruning − the tree is pruned by halting its construction early.
 Post-pruning - This approach removes a sub-tree from a fully grown tree.

39
STEPS: (Using Weka Explorer)
1.Open Weka tool and choose Explorer.
2.Click - Open file… in preprocess tab –choose weather.nominal.arff.
3.Goto Classify tab – click choose button – selelct J48.

40
Click – Start button.
Right click trees-J48 – choose Visualize Tree option.
OUTPUT :
Decision Tree

41
STEPS: (For Knowledge Flow)
Click Knowledge Flow in Weka GUI Chooser.
Draw ArrfLoader and select the filename.
To create the following and click Run button and Right click the TextViwer and select Show
Result.

42
Right click GraphViewer – click Show Plots.
OUTPUT:
Decision Tree

43
8. LAB SESSION 7: SUPPORT VECTOR MACHINE
AIM:
This experiment illustrates the use of Support vector classifier in weka. The sample data set used
in this experiment is vote dataset available in arff format. This document assumes that appropriate
data preprocessing has been performed.
8.1.SUPPORT VECTOR MACHINES
Support Vector Machines (SVMs) are supervised learning methods used for classification and
regression tasks that originated from statistical learning theory as a classification method, SVM
is a global classification model that generates non-overlapping partitions and usually employs all
attributes. The entity space is partitioned in a single pass, so that flat and linear partitions are
generated. SVMs are based on maximum margin linear discriminates, and are similar to
probabilistic approaches, but do not consider the dependencies among attributes.
SVM—History and Applications
 Vapnik and colleagues (1992)—groundwork from Vapnik & Chervonenkis‘statistical
learning theory in 1960s.
 Features: training can be slow but accuracy is high owing to their ability to model
complex nonlinear decision boundaries (margin maximization).
 Used both for classification and prediction.
 Applications: handwritten digit recognition, object recognition, speaker identification,
benchmarking time-series prediction tests.
Algorithm
 Define an optimal hyper plane: maximize margin
 Extend the above definition for non-linearly separable problems: have a penalty term for
misclassifications.
 Map data to high dimensional space where it is easier to classify with linear decision
surfaces: reformulate problem so that data is mapped implicitly to this space.
PROCEDURE:
1. We begin the experiment by loading the data (vote.arff) into weka.
2. Next we select the classify tab and click choose function button to select the Support vector
machine (SMO).
3. Now we specify the various parameters. These can be specified by clicking in the text box to
the right of the chose button.
4. Under the “text “options in the main panel. We select the 10-fold cross validation as our
evaluation approach. Since we don’t have separate evaluation data set, this is necessary to get

44
a reasonable idea of accuracy of generated model.
5. We now click ”start” to generate the model .the ASCII version of the tree as well as evaluation
statistic will appear in the right panel when the model construction is complete.
6. Note that the classification accuracy of model is about 69%.this indicates that we may find
more work. (Either in preprocessing or in selecting current parameters for the classification)
7. The run information of the support vector classifier will be displayed with the correctly and
incorrectly classified instances.
STEPS:
Open Weka tool – click Explorer in Weka GUI Chooser.
In preprocess tab click Open file.. – choose file name (vote.arff).

45
Goto Classify tab - click Choose button – select SMO option.
Click Start button.

47
9. LAB SESSION 8: APPLICATIONS OF CLASSIFICATION FOR WEB
MINING
AIM
To analyze an application using weka tool.
9.1 WEB MINING
Use of data mining techniques to automatically discover interesting and potentially useful
information from Web documents and services.
Web mining may be divided into three categories.
 Web content mining
 Web structure mining
 Web usage mining
Web mining is the application of data mining techniques to extract knowledge from web data, i.e.
web content, web structure, and web usage data.
9.1.1. WEB CONTENT MINING
Web content mining is the process of extracting useful information from the contents of web
documents. Content data is the collection of facts a web page is designed to contain. It may
consist of text, images, audio, video, or structured records such as lists and tables. Application of
text mining to web content has been the most widely researched. Issues addressed in text mining
include topic discovery and tracking, extracting association patterns, clustering of web
documents and classification of web pages. Research activities on this topic have drawn heavily
on techniques developed in other disciplines such as Information Retrieval (IR) and Natural
Language Processing (NLP). While there exists a significant body of work in extracting knowledge
from images in the fields of image processing and computer vision, the application of these
techniques to web content mining has been limited.
9.1.2. WEB STRUCTURE MINING
The structure of a typical web graph consists of web pages as nodes, and hyperlinks as edges
connecting related pages. Web structure mining is the process of discovering structure
information from the web. This can be further divided into two kinds based on the kind of
structure information used.
Hyperlinks
A hyperlink is a structural unit that connects a location in a web page to a different location,
either within the same web page or on a different web page. A hyperlink that connects to a
different part of the same page is called an intra-document hyperlink, and a hyperlink that
connects two different pages is called an inter-document hyperlink. There has been a significant
body of work on hyperlink analysis, of which provide an up-to-date survey.
Document structure
In addition, the content within a Web page can also be organized in a tree- structured format,
based on the various HTML and XML tags within the page. Mining efforts here have focused on
automatically extracting document object model (DOM) structures out of documents.
9.1.3. WEB USAGE MINING
Web usage mining is the application of data mining techniques to discover interesting usage
patterns from web usage data, in order to understand and better serve the needs of web-based

48
applications Usage data captures the identity or origin of web users along with their browsing
behavior at a web site. Web usage mining itself can be classified further depending on the kind of
usage data considered.
Web server data
User logs are collected by the web server and typically include IP address, page reference and
access time.
Application Server Data
Commercial application servers such as Weblogic, StoryServer, have significant features to
enable E-commerce applications to be built on top of them with little effort. A key feature is the
ability to track various kinds of business events and log them in application server logs.
Application Level Data
New kinds of events can be defined in an application, and logging can be turned on for them
generating histories of these events. It must be noted however, that many end applications require
a combination of one or more of the techniques applied in the above the categories.

49
10. LAB SESSION 9: CASE STUDY ON TEXT MINING
AIM:
To perform the text mining using Weka tool.What is text mining?
 Data mining in text: find something useful and surprising from a text collection.
 Text mining vs. information retrieval.
 Data mining vs. database queries.
Types of text mining
 Keyword (or term) based association analysis.
 Automatic document (topic) classification similarity detection.
 Cluster documents by a common author.
 Cluster documents containing information from a common source.
 Sequence analysis: predicting a recurring event, discovering trends.
 Anomaly detection: find information that violates usual patterns.
 Discovery of frequent phrases.
 Text segmentation (into logical chunks).
 Event detection and tracking.
Information Retrieval
Information retrieval deals with the retrieval of information from a large number of text- based
documents. Some of the database systems are not usually present in information retrieval systems
because both handle different kinds of data. Examples of information retrieval system include.
 Online Library catalogue system.
 Online Document Management Systems.
 Web Search Systems etc.
The main problem in an information retrieval system is to locate relevant documents in a
document collection based on a user's query. This kind of user's query consists of some keywords
describing an information need.
In such search problems, the user takes an initiative to pull relevant information out from a
collection. This is appropriate when the user has ad-hoc information need, i.e., a short-term need.
But if the user has a long-term information need, then the retrieval system can also take an initiative
to push any newly arrived information item to the user.
This kind of access to information is called Information Filtering. And the corresponding
systems are known as Filtering Systems or Recommender Systems.
Basic Measures for Text Retrieval
We need to check the accuracy of a system when it retrieves a number of documents on the
basis of user's input. Let the set of documents relevant to a query be denoted as {Relevant}
and the set of retrieved document as {Retrieved}. The set of documents that are relevant and
retrieved can be denoted as {Relevant} ∩ {Retrieved}. This can be shown in the form of a Venn
diagram as follows

50
There are three fundamental measures for assessing the quality of text retrieval
 Precision
 Recall
 F-score
PRECISION
Precision is the percentage of retrieved documents that are in fact relevant to the query. Precision
can be defined as
Precision= |{Relevant} ∩ {Retrieved}| / |{Retrieved}|
RECALL
Recall is the percentage of documents that are relevant to the query and were in fact retrieved.
Recall is defined as
Recall = |{Relevant} ∩ {Retrieved}| / |{Relevant}|
F-SCORE
F-score is the commonly used trade-off. The information retrieval system often needs to trade-off for
precision or vice versa. F-score is defined as harmonic mean of recall or precision as follows
F-score = recall x precision / (recall + precision) / 2
SAMPLE EXERCISE
Spam.arff

51
Open the file spam.arff
Choose Filter -> unsupervised->attributebased->StringToWordVector
Choose StringToWordVector
Choose the attributes for filter

52
Click Start button.
Choose Edit -> Select Attribute as class
Choose Classify ->Naïve Bayes

Bahir Dar Data Mining Lab-Weka Edited.pdf

More Related Content

Similar to Bahir Dar Data Mining Lab-Weka Edited.pdf

More from amare lakew

Recently uploaded

Bahir Dar Data Mining Lab-Weka Edited.pdf