Morphological Structure Discovery

Part 1:

After using undivide++ on the cleaned corpus the ouput obtained is here.

Part 2:

The hand annotated list is here.

Part 3:

The results obtained are as follows

Hits(H) = 223

Insertion(I) = 284

Deletion(D) = 20

Precision = 43.98%

Recall = 91.76%

F-score = 59.46%

Code:

Code used can be downloaded here.

Part 4: Work to be done in group

Here are the files groundtruth-morph.txt and problems-morph.txt