3. Malaga Analysis

By using the program coma and the Malaga library the wordlist of a corpus may be morphologically analysed. The analysis are written to the following files:

• malaga_id.mat // type mapping to analysis
• malaga_id.man // stored analyses
• malaga_id.sym_c // malaga symbol table

After the complete analysis, a statistic profile is written to standard out: (example)

Malaga morphological analysis for corpus limas:
Consumed time: 1154.78 sec
Number of types: 121650
Number of recognized types: 100051 (82.245%)
Number of unknown types: 21599
Number of tokens: 1236774
Number of recognized tokens: 1176201 (95.1023%)
Number of unknown tokens: 60573
Result files : 
limas.matest.mat
limas.matest.man
limas.matest.sym_c

A copy of the Malaga symbol associated with the analysis is made, thus enabling the use of the stored Malaga analysis independent of the rule files.

coma supports the following options:

• coma -apu corpus_id malaga_id pro-file
• -a analyse wordlist
• -p print analysed wordlist to standard out
• -u print unknown words to standard out