SPCDIS CHIP Role View Use Case 1

Printer-friendly version

Scenario

A user Alice is looking at a listing of CHIP Velocity images, similar to what is currently provided by the browsing interface at http://mlso.hao.ucar.edu/cgi-bin/mlso_data.cgi?2010&CHIP. In addition to displaying image listings, the browsing interface Alice is using provides links to provenance traces for each image. These provenance traces describe the steps taken by NCAR’s section of the CHIP pipeline (i.e., they trace from the raw image data provided by MLSO to the final data products published by NCAR).

In this use case, the objective of the provenance is to indicate:

Which processes did NCAR’s section of the CHIP Pipeline perform to generate this Velocity image?

How should the answer to this question vary for:

  • Alice the ACOS staff member?
  • Alice the outside domain expert?
  • Alice the citizen scientist.

Definition of roles

From Stephan Zednik:

    

Pipeline Scientist/Engineer:
Users who know the ins and outs of the processing pipeline and are likely to use provenance for quality control and processing verification. These users also build new highly-specialized products from the data products and will ask very specific and low-level questions of the processing such as "What collaboration processes (and their versions) were being run 11 years ago?", or "When was the last time filter x was replaced?".

This user will be able to see and understand everything, but an organized presentation of the information will still be important for navigation.

Examples: Joan, Leonard, and Don at HAO.

Expert Consumer / Research Scientist:
Scientist in the field with a background in the science of the data product but is likely not intimately familiar with the processing applied in the pipeline. This user will likely be highly curious of what processing (descriptions, explanations, how it relates to the science) is being applied. This user is looking to build a level of trust towards what would normally be a black-box system.

This user will not have as deep an understanding of all elements in the provenance - but will be looking at explicit, detailed information to build trust and understanding.

Examples: Domain experts not at HAO who consume the data products.

Non-expert Consumer:
Users who may be viewing the provenance or explanation as a general learning activity. This user likely has little background knowledge on the science or processing related to the data product and is therefore likely to implicitly trust the product and use the explanation to build general understanding. High levels of abstraction, general and educational descriptions, etc are important for this level of user.

Examples: Independent scientists, high school teachers, etc.