README

The code here is the one used in the article "Continual General Chunking Problem and SyncMap" which can be accessed by the link below:
https://arxiv.org/pdf/2006.07853.pdf
Please refer to the article for details on the SyncMap and the continual chunking problem.

========
keras==2.13.1
tensorflow==2.13.0
networkx==3.1
numpy==1.23.5
matplotlib==3.3.4
scikit-learn==1.3.2
scipy==1.10.1
========

First create a directory with the name "output_files" in the current directory.

Then run the following:
$python main.py problem_number output_filename

where, output_filename is the name of the output file and problem_number is the number of the problem.
Respectively, 
1: Mixed Structures
2: Fixed Chunks 
3: Sequence 2
4: Sequence 1
5: Long Chunks
6: Overlap 1
7: Overlap 2

For continual versions run the following:
$python dynamic_main.py problem_number output_filename

problem_number refers to the problems below:

1: Continual Mixed
2: Continual Fixed


The default learning algorithm is set for SyncMap.
For changing the method, uncomment the code between the following lines.
VAE refers to the modified Word2Vec mentioned in the article.
###### VAE #####
###### VAE #####