Page 118 - Big data - Concept and application for telecommunications

3 Big data - Concept and application for telecommunications

Figure 7-11 – Configuration of logical components for big data provenance
The logical components shown in Figure 7-11 are as follows:

– provenance model management. This logical component manages provenance information
compatibility among different BDSPs. This logical component validates the provenance information
transmitted from outside based on the big data provenance model (see clause 7.2). The valid
provenance information is then encoded as a common model and delivered to provenance lifecycle
management component to store it;
– provenance lifecycle management. This component performs the recording and deletion of
provenance information according to store, update and delete data (see clauses 7.3.1 and 7.3.2).
This logical component supports retrieving provenance information (see clause 7.3.3);
– analysis support. This logical component extracts the workflows from provenance information, and
stores them. From the stored workflows, this logical component retrieves the candidate analysis
workflows based on the information of BDSP's data analysis functions and data. For the request of
provenance information or workflow from the different system (e.g., external BDSP), this logical
component may check the adaptability of the computational environment, and map to an
equivalent function for that system. This logical component also supports automating data analysis
process based on update of data, adding user annotation on provenance information, and managing
the relationship between BDSP's functions and data;
NOTE – Based on the relationship between functions and data in provenance information, it is
possible to query the list of available data with functions, and the list of functions applicable to the
data.
– provenance sharing policy management. This logical component manages multiple sharing policies
on provenance information. When exporting a provenance information, a BDSP checks the sharing
policy and may simplify it before sending to another BDSP;
– personally identifiable information (PII) management. This logical component checks whether data
instance contains PII when recording a provenance unit. This logical component also requests a
protection mechanism to BDSP on provenance information that includes PII;
– monitoring. This logical component monitors changes in value about computational environment
and responsible party in provenance information. When changes are detected, this logical
component updates them.

8 Functional requirements of big data provenance

8.1 Provenance lifecycle requirements

Provenance lifecycle requirements include:
– (provenance model description) It is required that BDSP supports the model for big data
provenance information;

110 Static data – Data provenance, data formats and trust