My research

My information and CV may also interest you.

Governance rules modeling

Terms and Conditions widely exist, and are usually written in natural languages. However, people tend to ignore them because of the complexity or large volume. The situation is the same for scientific research, especially for research data.
We seek to solve this problem by modeling these terms and conditions (aka. data governance rules) using a logic form understood by machines and humans.
The rules will propagate along with the data, and be checked accordingly.

The two main objectives are:

  1. Model data governance rules (aka. Terms and Conditions, in the field of research data)
  2. Present a reasoning system over data-flow information (e.g. provenance)

Our solution, Dr.Aid (Data Rule Aid), differs from other research in:

  1. Dr.Aid supports obligations, unlike other research focusing on access controls
  2. Dr.Aid deals with multi-input-multi-output (MIMO) processes in multi-staged graphs, unlike other research focusing on linear / pipeline-style processing or a central server
  3. Dr.Aid can derive the data-use rules for the output data based on the processing, so reasoning for further users works automatically
  4. Dr.Aid's language model builds on top of attribute, an abstract construct. It can be extended to support other semantics (e.g. prohibitions, etc), out side of the current implemented semantics (obligations), withouth modifying the main part of the language features.

See 5-page slides for a summary on the current research results.

See the doc archives for more documents / papers describing my work.

links

social