Page 53 - Big data - Concept and application for telecommunications

Big data - Concept and application for telecommunications 1

NOTE 2 – Clustering analysis: This supports k – means, k – center point, overlapping clustering, fuzzy
clustering, etc., to classify data into different classes or clusters according to their similarity.
NOTE 3 – Association analysis: This supports some specific algorithms to find associations between
stored data. Examples of association algorithms include Apriori algorithm and Frequent Pattern
Growth algorithm. Apriori algorithm and Frequent Pattern Growth algorithm are two classical
association analysis algorithms which can mine the associations through the frequency of data
appearing together in the dataset.
NOTE 4 – Regression analysis: This supports linear regression and logistic regression and other
algorithms, for estimating the relationships among data.
NOTE 5 – Customization of analysis supports the customization of detail data analysis methods
according to a customer's specific requirements.
– setting up procedures which enable the analysis using registered analysis methods in the analysis
function registry;
– executing analysis process according to the procedures.
7.1.5 Data storage functional component

The data storage functional component is responsible for storing data. This functional component also
provides different types of storage for different data types and different database types while storing data.

This functional component provides:
– provisioning storage considering the various types of data storage, database, and different types of
data such as structured data, unstructured data, and semi-structured data;
NOTE 1 – Data storage types include block storage, file storage and object storage.
NOTE 2 – Databases include Relation database, No SQL database.
NOTE 3 – Unstructured data can include mass data, such as log files, video, audio data, email, Web
pages, data generated on social-media sites. Semi-structured data can include data stored in XML,
HTML and other format documents. Structured data can include record data persistent in databases
(see [ITU-T Y.3600]).
– allocating the appropriate storage when a storage usage request is initiated;

– releasing storage when the storage usage is terminated;
NOTE 4 – The data storage functional component interworks with the data collection functional
component (see clause 7.1.1) to identify the characteristics of the data such as data type, data
volume and so on.
– storing data on various storage systems. It supports storage mirroring and provides data
fragmentation to distribute and store data on distributed storage systems. This provides the ability
to update data;
NOTE 5 – Distributed storage system stores data on multiple independent storages. It adopts the
scalable system structure, and uses multiple storage servers which are used to share the storage
load.

NOTE 6 – Storage mirroring is the replication of logical storage volumes onto separate physical disks.
– data indexing, stored together with data, to improve the speed of data retrieval operations.

7.2 Resource layer functional components
The resource abstraction and control functional component, in the resource layer functional components, is
extended for BDaaS (see Figure 7-3) with the following functional components:
– distributed processing functional component (see clause 7.2.1).

Basics of Big data 45