def get_top(q, docs, mod, km_mod, cl_as):
res = []
q_title = mod[q["desc"]]
q_vec = np.asarray(make_vector(q_title)).reshape(1, -1)
q_cluster = km_mod.predict(q_vec)
test_doc_ids = cl_as[q_cluster[0]]
for d_id in q["results"]:
if d_id not in test_doc_ids:
test_doc_ids.append(d_id)
for i in test_doc_ids:
doc = docs[i]
try:
doc_title = mod[doc["abstract"]]
except:
doc_title = mod[doc["title"]]
sim_score = utils.compute_sim(q_title, doc_title)
if i in q["results"]:
res.append(((i, q["results"][i], sim_score)))
else:
res.append(((i, 0, sim_score)))
for i in q["results"]:
doc = docs[i]
try:
doc_title = mod[doc["abstract"]]
except:
doc_title = mod[doc["title"]]
sim_score = utils.compute_sim(q_title, doc_title)
if i not in test_doc_ids:
res.append(((i, q["results"][i], 0)))
return res
def get_top(q, docs, mod, km_mod, cl_as):
res = []
q_title = mod[q['desc']]
q_vec = np.asarray(make_vector(q_title)).reshape(1, -1)
q_cluster = km_mod.predict(q_vec)
test_doc_ids = cl_as[q_cluster[0]]
for d_id in q['results']:
if d_id not in test_doc_ids:
test_doc_ids.append(d_id)
for i in test_doc_ids:
doc = docs[i]
try:
doc_title = mod[doc['abstract']]
except:
doc_title = mod[doc['title']]
sim_score = utils.compute_sim(q_title, doc_title)
if i in q['results']:
res.append(((i, q['results'][i], sim_score)))
else:
res.append(((i, 0, sim_score)))
for i in q['results']:
doc = docs[i]
try:
doc_title = mod[doc['abstract']]
except:
doc_title = mod[doc['title']]
sim_score = utils.compute_sim(q_title, doc_title)
if i not in test_doc_ids:
res.append(((i, q['results'][i], 0)))
return res
