DATA can be described into three categories. 1. Call

DATA
MINING IN TELECOMMUNICATIONS INDUSTRY

The telecommunications industry adopted
data mining technology at a very early stage and thus there are its many data
mining applications. The main reason of this is because telecommunication
companies  generate and store large
amounts of high-quality data, have a very big customer base, and works in a
rapidly changing and largely competitive environment.

We Will Write a Custom Essay Specifically
For You For Only $13.90/page!


order now

Telcommunications generate a large
amount of data which is classified as such as call detail data, network data , customer
data. Call detail data describes the calls over the telecommunication networks
while network data tells the status of  hardware and software components in a network.
Customer data describes the telecommunication customers. Such rich data is an arable
environment for number of data mining applications which is built with the goal
of minimizing some of the most critical business issues in telecommunications.

 

Telecommunication companies use data
mining techniques enhance marketing efforts, detect fraud, and can manage their
telecommunication networks in a more better way. Although these companies faces
numerous data mining challenges due to large size of data sets.

 

TYPES OF TELECOMMUNICATION DATA

 The first step in the data mining
process is to understand the data. Without proper knowledge of data it is not useful
to develop application. In Telecommunications data can be described into three
categories.

 

1. Call Detail Data

Every
time when a call is made on a telecommunications network, detailed information
of the call is stored in call detail record. Call detail records contains
plentiful of  information to describe the
important characteristics of each call. At a minimum, each call detail record
will contain the starting and ending phone numbers, the date and time of the
call and the duration of the call. Call detail records are made in real time and
thus will be available directly for data mining. This is in contrast with billing
data, which is hardly made available only once per month. Call detail records
are not used directly for data mining since the purpose of data mining
applications is to extract knowledge at customer level, not at the level of
individual phone calls. Thus, the call detail records related with a customer
must be summed up into a single record which describes the customer’s calling
behavior.

 

 

2. Network Data

Telecommunication
networks are complex combinations of equipment which comprised of thousands of
interconnected components. Each network element is able of generating error and
status messages, which leads to a large amount of network data. This data must
be kept and analyzed in order to support network management functions, such as
fault isolation. Network data is also generated in real time as a data stream
and must often be summed up in order to be useful for data mining

 

3. Customer Data

Telecommunication
companies like number of large businesses, may have numbers of customers.
Essencially this means maintaining a database of information of these
customers. This information will include address and name information and may
consist other information such as service plan and contract information, family
income, credit score and payment history. This information may be added with
data from external sources, such as from credit reporting agencies.

 

As Data is the base of Telecommunication so data
mining is used to carry operations on data and get the desired results. Numerous data mining applications
have been developed in the telecommunications industry. However, most of them
fall into one of the following three categories:

Marketing,
Fraud Detection, and Network Fault Isolation and prediction

FRAUD
DETECTION

Fraud
is a serious problem for telecommunication companies which leads to billions of
dollars in lost revenue each year. Fraud can be classified into two categories:
subscription fraud and superimposition fraud. “Subscription fraud happens when
a customer creates an account with the motive of never paying for the account
charges”. “Superimposition fraud involves a legitimate account with some
legitimate activity, but also includes some superimposed illegitimate activity
by a person other than the account holder”. Superimposition fraud presents a
larger problem for the telecommunications industry and for this reason we focus
on applications for identifying this type of fraud.

 

The most common method for detecting
superimposition fraud is to compare the customer’s current calling behavior
with a profile of his past usage, using

deviation detection and anomaly
detection methods. The profile must be cabable to be fastly updated because of
the amount of call detail records and the need to detect fraud in a timely
manner. “Cortes and Pregibon” generated a signature from a data stream of
call-detail records to brively describe the calling behavior of customers and
then anomaly detection is used to “measure the unusualness of a new call relative
to a particular account.”

 

MARKETING OR CUSTOMER PROFILING

Telecommunication
companies keeps numerous of data about their customers. In addition to the
general customer data that most businesses collect, telecommunication companies
also maintains call detail records, which accurately details the calling
pattern of each customer. This information can be used to profile the customers
and the profiles can then be used for marketing and for forecasting.

The main element of modeling a
telecommunication customer’s value is measuring how long they will remain with
their current carrier. This issueis of interest in its own right since if a
company can tell when a customer is likely to leave, it can take

proactive steps to stop the
customer. “The process of a customer leaving a company is referred to as churn”. Churn analysis includes
building a model of

customer attrition.

 

NETWORK
FAULT ISOLATION

 

 As told  above Telecommunication networks are of extremely
complicated configuration and they contains many interconnected components.
These network components generate status and alarm messages every time. So in
order to identify the network faults the alarms should be analyzed
automatically. So the data mining helps to automatically analyze and identify
the faults so that they can be resolved as faster as can.