Automated Terminology Harmonization, Extraction and Normalization for Analytics - ATHENA
Preliminary analysis
Application Design
Sidebar
documentation:athena:problem_description
Table of Contents
Problem description
Two kinds of changes, which will require source adaptation, can be distinguished.
Changes in the structure
- New column added.
- Column was removed.
- Column data type changed.
- Column name changed.
- New relation appeared.
- Relation is removed.
- Relation type changed.
Change of data
- Records added.
- Records removed.
- Record altered.
Versioning should be kept in mind also. For example, we have “The Vocabulary” v.1 from 01.01.2012, and it has the records:
| Procedure code | Procedure name |
| А15 | Complete blood count |
| B17 | X-ray photograph |
Then we receive a new version of the “The Vocabulary” v.2 from 02.02.2012. It has the records:
| Procedure code | Procedure name |
| А15 | Ultrasonography |
| B17 | X-ray photograph |
Having the procedures account with structure:
| Person_id | Procedure code |
| P01 | A15 |
We will encounter the invalid data after “The Vocabulary” v.2 will be applied.
documentation/athena/problem_description.txt · Last modified: 2015/03/18 10:29 by gleb_malikov
