Loading README +3 −0 Original line number Diff line number Diff line Loading @@ -7,3 +7,6 @@ In my approach, for now the morphological rules are crafted by hand (explicity f python3 main.py -h Some usecases: TODO +5 −0 Original line number Diff line number Diff line Loading @@ -6,6 +6,11 @@ ===Macherey Splitter=== - Improve the segment mode (output split should be reversible) - Genaueres Untersuchen von Himmelköper/Himmelskörper. Evtl Kosten für Regelanwendung verringern? - Kosten mal an eine Beispiel von Hand durchrechnen? ===Zusätzliche Unigram Filter=== ===Things to do in the future=== Loading eval/eval.py +11 −2 Original line number Diff line number Diff line Loading @@ -11,10 +11,13 @@ args = parser.parse_args() def matching(parts, parts_gold): for w, w_gold in zip(parts, parts_gold): w_gold_options = w_gold.split("|") match = False for option in w_gold_options: if w==option: return True match = True if not match: return False return True if args.output_file: correct_splits = 0 Loading Loading @@ -75,6 +78,12 @@ if args.output_file: wrong = wrong_faulty_splits + wrong_splits + wrong_non_splits accuracy = correct/ (correct + wrong) print ("correct_splits:", correct_splits, sep=" ") print ("correct_non_splits:", correct_non_splits, sep=" ") print ("wrong_non_splits:", wrong_non_splits, sep=" ") print ("wrong_faulty_splits:", wrong_faulty_splits, sep=" ") print ("wrong_splits", wrong_splits, sep=" ") print ("--------------") print ("Precision:", precision, sep= " ") print ("Recall:", recall, sep= " ") Loading Loading
README +3 −0 Original line number Diff line number Diff line Loading @@ -7,3 +7,6 @@ In my approach, for now the morphological rules are crafted by hand (explicity f python3 main.py -h Some usecases:
TODO +5 −0 Original line number Diff line number Diff line Loading @@ -6,6 +6,11 @@ ===Macherey Splitter=== - Improve the segment mode (output split should be reversible) - Genaueres Untersuchen von Himmelköper/Himmelskörper. Evtl Kosten für Regelanwendung verringern? - Kosten mal an eine Beispiel von Hand durchrechnen? ===Zusätzliche Unigram Filter=== ===Things to do in the future=== Loading
eval/eval.py +11 −2 Original line number Diff line number Diff line Loading @@ -11,10 +11,13 @@ args = parser.parse_args() def matching(parts, parts_gold): for w, w_gold in zip(parts, parts_gold): w_gold_options = w_gold.split("|") match = False for option in w_gold_options: if w==option: return True match = True if not match: return False return True if args.output_file: correct_splits = 0 Loading Loading @@ -75,6 +78,12 @@ if args.output_file: wrong = wrong_faulty_splits + wrong_splits + wrong_non_splits accuracy = correct/ (correct + wrong) print ("correct_splits:", correct_splits, sep=" ") print ("correct_non_splits:", correct_non_splits, sep=" ") print ("wrong_non_splits:", wrong_non_splits, sep=" ") print ("wrong_faulty_splits:", wrong_faulty_splits, sep=" ") print ("wrong_splits", wrong_splits, sep=" ") print ("--------------") print ("Precision:", precision, sep= " ") print ("Recall:", recall, sep= " ") Loading