This project annotates a .conllup file with a given terminology in the Romanian language.
This is done by running the annotate.py script. For instance, using the files found in examples, one can annotate a .connlup files as follows:
python annotate.py --data_path examples/input.conllup --output_path examples/output.conllup --terminology_path examples/terminology.csv
The data_path parameter points to the .conllup file to be annotated and output_path is the resulted .conllup file. The terminology_path parameter contains the
terminology to be used.
One can also change the lemmas used by this script by changing the lemma_path parameter (defaults to examples/tbl.wordform.ro) to point to another file. Also, the maximum number of words that can exist in a terminology can be adjusted using the max_terminology_words paramter (defaults to 10).
In addition, because a terminology can contain multiple languages, one can select the desired langauge using the terminology_langauge paramter (defaults to ro).
Important Note: Please follow the format in the terminology file, so the script knows how to parse it. This implies using the same headers, separated by |, as
exemplified in examples/terminology.csv:
E_ID|L_CODE|T_TERM|T_TYPE|
11111|ro|Paris|Term|
22222|ro|statului portului|Term|
33333|ro|ordinul|Term|
44444|ro|Monitorul Oficial|Term|
55555|ro|Belgia|Term|
66666|en|CFR|Abbrev|
The terminology file can have any number of headers, but the three presented above must be present (i.e. E_ID, L_CODE, T_TERM, T_TYPE).