def essay_jaccard_similarity(kunci_jawaban,
jawaban,
char=False,
fitur=True,
praproses=True):
if type(kunci_jawaban) != list:
kunci_jawaban = [kunci_jawaban]
kunci_jawaban = cek_negasi_list(kunci_jawaban)
kunci_jawaban = stopword_list(kunci_jawaban)
kunci_jawaban_unik = " ".join(get_unik(kunci_jawaban))
# print(kunci_jawaban_unik)
if praproses == True:
jawaban = praproses_(jawaban, kunci_jawaban_unik)
# print(jawaban)
simm = list()
idx = list()
bobot = list()
for ix, kj in enumerate(kunci_jawaban):
fitur_ = list(set(kj.split()))
if char == True or fitur == True:
t = w.tf_idf(kj, jawaban, vocab=fitur_, fitur=fitur, char=char)[1]
t_x = t.transform([kj, jawaban]).toarray()
else:
t_x = w2.tf_idf([kj, jawaban], vocab=fitur_)
bobot.append(t_x)
# print(t_x)
jaccard = simi.jaccard(t_x[0], t_x[1])
simm.append(jaccard)
idx.append(ix)
max_sim = max(simm)
max_idx = simm.index(max_sim)
return [max_sim, max_idx, bobot, jawaban]
def essay_dice_similarity(kunci_jawaban, jawaban, bychar=False, fixed=False):
if type(kunci_jawaban) != list:
kunci_jawaban = [kunci_jawaban]
kunci_jawaban = cek_negasi_list(kunci_jawaban)
kunci_jawaban_unik = " ".join(get_unik(kunci_jawaban))
# print(kunci_jawaban_unik)
jawaban = praproses(jawaban, kunci_jawaban_unik)
# print(jawaban)
simm = list()
for kj in kunci_jawaban:
# print(t2c(kj))
t = w.tf_idf(kj, jawaban, fitur)
# print(t.A)
dice = simi.dice_similarity(t.A[0], t.A[1])
print(dice)
def essay_jaccard_similarity(kunci_jawaban, jawaban, bychar=False, fixed=False):
if type(kunci_jawaban) != list:
kunci_jawaban = [kunci_jawaban]
kunci_jawaban = cek_negasi_list(kunci_jawaban)
kunci_jawaban_unik = " ".join(get_unik(kunci_jawaban))
# print(kunci_jawaban_unik)
jawaban = praproses(jawaban, kunci_jawaban_unik)
# print(jawaban)
simm = list()
for kj in kunci_jawaban:
fitur = list(set(kj.split()))
#print(fitur)
t = w.tf_idf(kj, jawaban, fitur)
jaccard = simi.jaccard(t.A[0], t.A[1])
simm.append(jaccard)
return [max(simm), transform(max(simm))]
def essay_dice_similarity(kunci_jawaban, jawaban, char=False, fitur=False):
if type(kunci_jawaban) != list:
kunci_jawaban = [kunci_jawaban]
kunci_jawaban = cek_negasi_list(kunci_jawaban)
kunci_jawaban_unik = " ".join(get_unik(kunci_jawaban))
# print(kunci_jawaban_unik)
jawaban = praproses(jawaban, kunci_jawaban_unik)
# print(jawaban)
simm = list()
for kj in kunci_jawaban:
fitur_ = list(set(kj.split()))
t = w.tf_idf(kj, jawaban, vocab=fitur_, fitur=fitur, char=char)
# print(t.A)
dice = simi.dice_similarity(t.A[0], t.A[1])
simm.append(dice)
return [round(max(simm), 10), transform(max(simm))]
def essay_cosine_similarity(kunci_jawaban, jawaban, char=False, fitur=True):
if type(kunci_jawaban) != list:
kunci_jawaban = [kunci_jawaban]
kunci_jawaban = cek_negasi_list(kunci_jawaban)
kunci_jawaban_unik = " ".join(get_unik(kunci_jawaban))
# print(kunci_jawaban_unik)
jawaban = praproses(jawaban, kunci_jawaban_unik)
# print(jawaban)
simm = list()
for kj in kunci_jawaban:
if char == True and fitur == True:
print("tt", end="")
elif char == True and fitur == False:
print("tf", end="")
elif char == False and fitur == True:
print("ft", end="")
elif char == False and fitur == False:
print("ff", end="")
fitur_ = list(set(kj.split()))
t = w.tf_idf(kj, jawaban, vocab=fitur_, fitur=fitur, char=char)
print(t.A)
cosine = simi.cosine_similarity(t.A[0], t.A[1])
simm.append(cosine)
return [round(max(simm), 10), transform(max(simm))]