Page 84 - Big data - Concept and application for telecommunications

2 Big data - Concept and application for telecommunications 2

8.3 Requirements for data delivery

Data delivery requirements include:
1) it is required that BDSP provides effective data transmission methods;
NOTE 1 – Effective data transmission methods include secure connection control, network
consistency & scalability, and weight lightening (e.g., data compression), etc.
2) it is required that BDSP requests a retransmission when the data transmission fails;
3) it is recommended that BDSP supports various data delivery methods;
NOTE 2 – Data delivery method includes download at once, online cloud storage service, a form of
database (e.g., Structured Query Language interface), an event push, etc.
4) it is required that DP:data broker mediates the secure transmission mechanism between the data
provider and the big data service provider;
NOTE 3 – An example of mediation of secure transmission is to provide public key for data provider
and big data service provider.
5) it is recommended that BDSP supports the combination, aggregation, integration and
intermediation of datasets from multiple data sources;
6) it is recommended that BDSP assembles data according to published data specifications;
NOTE 4 – A data specification includes information about real-time data material with data access
information, collection information, and information about a form in which data are stored, etc. In
this case, BDSP connects to DP:data supplier. The BDSP requests data from DP:data supplier and
stores the received data in order to perform data assembling.
7) it is required that DP:data supplier provides mechanisms to protect integrity, availability and
authenticity of data;
8) it is recommended that DP:data broker provides charging methods on the data.

8.4 Requirements for customer support

Customer support requirements include:
1) it is required that BDSP provide interfaces to BDC for operating a workflow and data analysis;
NOTE 1 – Workflow is the series of analysing events that are necessary to complete a task. BDC can
utilize the workflow for automating processes related to the analysis of big data.
2) it is recommended that BDSP supports a method for utilizing a user defined analysis algorithm;
NOTE 2 – A method for utilizing a user defined analysis algorithm includes algorithm development
functions, a common model to describe analysis algorithm, and an encoding tool for converting
them to a common model, etc.
3) it is recommended that BDSP provides a meta-information for an analysed result to BDC.
NOTE 3 – A meta-information may include metadata of analysed data, and functions with
parameters which had been applied.

8.5 Requirements for data quality management
Data quality management requirements include:

1) it is recommended that BDSP provides an evaluation method for managing data quality;
NOTE – Evaluation methods includes scoring (rating), reviewing, etc.
2) it is recommended that BDSP provides evaluation result about data quality;
3) it is recommended that DP:data broker is able to access that the quality of data provided by the
DP:data supplier is reliable.

76 Moving data – Data exchange and data flow