This is source code of project "A Hybrid Approach to Learning DL Ontology From Text"
- Metamap version 2013. When install, put
public_mm
folder in the same directory with this source code directory - Python. To test, run
python generatePositiveExamplesOriTest.py
- Stanford Parser. After extracting the folder, make sure that you rename the outer folder to become
stanford-parser
(renamestanford-parser-full-20xx-xx-xx
tostanford-parser
). Then put that folder in the same directory with this source code directory
- Compile all
.java
files by runningjavac *.java
- Run GUI class by command
java GUI
- Insert input file containing biomedical text. You can try to insert file
input.in
in this folder - Click
Process
and see the result
User can see a set of sentences where a GCI comes from by clicking the ?
symbol.
In the second box, the system will show those set of sentences.
User may change the role relation between two concept names.
- To validate, click the
V
symbol. - In third box, system shows two concept names and their role in the format
<first concept name> | <role relation> | <second concept name>
- User can change role relation by writing this abbreviation:
AM
: Associated MorphologyAM-1
: Associated Morphology (inverse role)FS
: Finding SiteFS-1
: Finding Site (inverse role)CA
: Causative AgentCA-1
: Causative Agent (inverse role)IS
: Is-A relationIS-1
: Is-A relation (inverse role)
- and finally, click
Train
.