Page 160 - Big data - Concept and application for telecommunications

4 Big data - Concept and application for telecommunications

5 Conventions

None.

6 Overview of big data analysis in mobile Internet services
Due to improvements in the computation and storage ability of smart devices and the enhanced transmission
rate in telecommunication networks, as well as the advent of service platforms, the mobile Internet services
are more popular and more widely used. Due to frequent interaction between the users, networks, multiple
types of devices, and service providers, data is growing at an unprecedented rate for a broad range of mobile
Internet services areas.

Generally, big data technologies provide two benefits, cost efficiency and new insights from the large data
volumes.

– Cost efficiency: Big data technologies were designed to offer a fast and scalable data process, for
example, the ‘click stream' analysis of user's activities. Web portal and more generally digital service
providers have a large volume of log data. Traditionally, they create a database and data warehouse
system in their data centres and perform analysis in the batch system using the relational database
management system (RDBMS). But the management cost has always been very high and there is no
way to extract new features from the data set. Therefore, many service providers have replaced or
are in the process of replacing existing systems to the distributed file system and pre-processing unit
(sub-system of big data). They can reduce the data management cost to half or less of what it used
to be.
– Extract new insight: Big data technologies enabled the development of analytics technologies which
move the analysis to a new degree from raw data to business insight. That solution now includes
prediction to the extent of allowing the implementation of decision assisting systems. This is indeed
very different from the conventional systems because the way the data processing logic handles the
raw data, and the kinds of information that can be extracted from the data set are already known.
The way big data technology was designed allows the delivery of analytics that, from the perspective
of a human being, would equate to guessing the result. For example, in many implementations, the
system helps the creation of "candidate answers" and tries to collect some evidence in order to
prove that one answer is better than others.

Whilst for mobile Internet services, cost efficiency is the most important benefit, getting new business insight
is equally, if not more, important for the next generation of mobile Internet services. As the data source is
not fixed and will be diverse, the analysis system could be used by malicious users or attackers to achieve
illegal or unethical purposes.
Mobile Internet services harvest data in their big data infrastructure from multiple sources and multiple data
dimensions with characteristics including, scale (volume), diversity (variety), high speed (velocity) and
possibly others like credibility (veracity) or business value. Big data analysis now drives nearly every aspect
of mobile Internet services to improve service quality and user experience. According to big data aggregation
and analysis, service providers can analyse users' interests more effectively and predict user's expectation
more accurately, thus, they can significantly improve and add value to their services, for example:
– mobile search application to precisely target users' search intentions in a timely way;
– mobile financial application to customize users' financial solutions in a timely way;
– mobile application recommendation to improve success rates of recommendations in a timely way.

Consequently, big data analysis, as a new technology, becomes a valuable business trend in the
telecommunication domain. Figure 1 shows big data analysis in mobile Internet services.

152 Security, privacy and data protection