Page 66 - ITU Journal - ICT Discoveries - Volume 1, No. 2, December 2018 - Second special issue on Data for Good

ITU JOURNAL: ICT Discoveries, Vol. 1(2), December 2018

metadata used in the scenario analysis and
implementations is provided in the form of a
pseudo code to describe the execution workflows.
4.1 Scenario analysis

A scenario analysis in a smart-home environment
is provided to validate the applicability of the
procedural metadata concept. This scenario
describes how metatags and procedural metadata
can be used for automatic electricity consumption
pattern estimation.
4.1.1 System environment

The system environment is designed to have data
storage (database), a convergence system (data
analyzer and organizer), and actors (smart home,
system services, and 3 party service providers).
rd
Connected devices in the smart home generate
data, which is transferred to a database and data
analyzer according to procedural metadata. The
Fig. 4 – Example of procedural metadata embedment data analyzer processes the metatags and their
values, with raw data if needed, to examines the
relations, rules, and policies as described in the
procedural metadata. Based on data analysis
results and procedural metadata, the data
organizer creates, updates, and manages the
metatags, then it makes decisions on the current
conditions. According to analyzed contexts, the
system service and 3 party service providers take
rd
proper actions with up-to-date descriptions of the
Fig. 5 – Example of security ensuring procedural metadata current situations.
4.1.2 Consumption pattern estimation scenario
needs to be already affiliated with and registered
in the secured and trusted domain, then it can For the electricity consumption pattern estimation
access the system according to its clear level. task, it is assumed that an advanced metering
Otherwise, that object must go through security infrastructure (AMI) device exists in the
and trust appreciations according to appropriate smart-home environment and that it periodically
procedural metadata. By denying unsecured or collects the electricity consumption data.
untrusted access, the data, devices, and systems
can ensure their security. With the collected data, the data analyzer performs
estimation tasks according to the procedural
4. PROOF OF CONCEPT metadata as follows:
– Retrieve pattern and peak time tag data for
The proof of concept is implemented in Matlab the past 7 days.
based on a scenario analysis in a smart-home
environment. The feasibility of the proposed – Organize the tag data into a sequence.
metadata to be applied to the web are shown as – Search the top 5 similar sequences in the da-
well. Self-defined metatags are used to ease the tabase.
implementation process. However, different types – Choose the pattern value of the day after the
of data formats, models, languages, vocabularies, end of each sequence as a candidate pattern.
ontologies, etc. can be used for procedural
metadata as long as they support a micro-
formatted tag structure. Hence, the procedural

44 © International Telecommunication Union, 2018