ABOUT THE SUBJECT

What domain ?

We have chosen the Telecommunications industry for that assignment.

For our group, Telecommunications was an obvious choice.

Indeed, this is such a large fields that we thought of that domain since the beginning.

Our propositions

To analyze the relation between the quality of the cover with the operator, the network and the location in France (association analysis), which could be useful on the customer side to choose an operator among others for instance.

To predict where the network or the quality of the cover will be the best in France (prediction), and so on the other hand to predict where the network would be the poorest so that investments can be made to improve it.

Market segmentation (clustering), that could be useful for commercial purposes, and increase the sales.

Example of stories and academic papers

Application of data mining in telecommunication industry in Nigeria

This article is about a data mining model in sales and marketing department of the Telecommunication industry in Nigeria.

This paper has been written in order to show the problem that faces the TI sales and marketing department, such as the inability in gaining precise view of targeted data, the inability to translate and formulate business question correctly, and the problem in addressing data quality.

The main purpose of this work is to develop and implement a data mining model which will be use in order to retain existing customers, manage and allocate resources more effectively.

The data mining techniques used for this work were classification, association, sequence discovery, visualization and prediction. The tools used to implement the model were PHP, JavaScript, CSS and HTML.

For this work, three products on sales and marketing department were considered :

SIM card
Airtime
Electronic recharge

The training data used range from 2008 to 2015 and was collected from the historical of one Telecommunication company (EMTs).

Pattern Mining in Telecom Data in Pakistan

This articles discuss about the trend analysis for predicting and finding interesting patterns on telephone calls data of Pakistan Telecommunication Company Limited.

In the past few years, interesting mining patterns has become a popular research issue.

Telecom industry want to invest in new technology for CRM (customer relationship management) in order to retain their customers, and thus to perform competitive analysis.

The main objective in this paper is to identify interesting trend in telephone calls by using statistical techniques, such as the ones previously discussed.

Those trends can be used in order to take decision in the future for instance.

The techniques used are autoregressive moving averages model, empirical analysis and also boxplot analysis.

The patterns found thanks to those techniques are really helpful for assisting decision making in the industry.

Data Mining helps Telcos create new revenue stream

The data mining and analytics techniques help Telecommunication companies to explore more revenue channels and monetize their existing assets.

The startup 3Loq is helping the mobile service providers to monetize their SMS assets thanks to analytics solution and data mining. The startup also help those companies to reach potential new customers.

The objective of the startup is to create a platform where the Telcos operators and the advertisers share their data.

The founder of this startup think that the data own by the companies on their customers could be really interessant for brands and retailers.

He also believe that SMS is the biggest channel for mobile marketing and can offer good revenue.

Using data mining to fight against fraud

In this case, they used Big data to fight fraud. They changed their approach to detect fraud, in fact the company’s big data-based solution now permits to detect fraud that wasn’t previously detectable.

With their new solution based on big data, they can access data more easily than before. They still generate more and more data, so it’s really hard to analyze in a real time, but in the same time they are innovating for Extreme scale with this approach.

The goal is to be able to know what happen every second, so they can react easily if they detect some fraud. The combination of Machine Learning and Petabyte System is the key to beat fraud definitely.

Google Sites

Report abuse