delivers services for executing
data analysis models in cloud environments and preserving
execution codes within existing problem solving environments
such as MATLAB [65], Octave [66] and R [67]. For each model
execution task, it first creates a Temporary Execution Directory
(TED). Then, it retrieves the required model files from Model
Repository and data through the Data Storage Layer, placing them
all in the TED. Next, it sets TED as the working directory and starts
executing the Main Function that is used as the entry point for
each model. The user declares the filename, inputs, outputs, and
actionable information of the Main Function during the model
publishing process. Within the Main Function, external functions
and subroutines can be called and embedded for performing
computations. Users and developers are able to configure the
integration and connections between data sources to the model,
such as the mappings of different data streams to the inputs and
outputs. The Main Function provides the ability for developers
to easily migrate their existing local data analysis functions and
subroutines to the Wiki-Health Analysis Framework. All the output
and intermediate files are kept in the TED. Finally, MEE conducts
the corresponding actions such as persisting output files and data
through Data Storage Layer to the underlying databases and trigger
message alerts to the user.