def run(q, d1, d2_s):
d2 = convert_string_to_dict(d2_s)
dimensions = return_keys(d1, d2)
v1 = mk_vector(d1, dimensions)
v2 = mk_vector(d2, dimensions)
v1_bin = binarise(v1)
v2_bin = binarise(v2)
coverage = sum(v1_bin * v2_bin) / len(q.split())
d1_vec = normalise(v1)
d2_vec = normalise(v2)
return cosine_similarity(d1_vec, d2_vec), coverage
def compute_pod_summary(name):
'''This function is very similar to 'self' in PeARS-pod'''
DS_vector = np.zeros(400)
word_vector = ""
freqs = {}
for u in db.session.query(Urls).filter_by(pod=name).all():
DS_vector += convert_to_array(u.vector)
for k, v in convert_string_to_dict(u.freqs).items():
if k in freqs:
freqs[k] += int(v)
else:
freqs[k] = int(v)
DS_vector = convert_to_string(normalise(DS_vector))
c = 0
for w in sorted(freqs, key=freqs.get, reverse=True):
word_vector += w + ':' + str(freqs[w]) + ' '
c += 1
if c == 300:
break
return DS_vector, word_vector
def compute_query_vectors(query):
""" Make distribution for query """
words = query.rstrip('\n').split()
# Only retain arguments which are in the distributional semantic space
vecs_to_add = []
words_for_freqs = []
for word in words:
words_for_freqs.append(word)
if word in dm_dict_en:
vecs_to_add.append(word)
vbase = np.array([])
# Add vectors together
if vecs_to_add:
# Take first word in vecs_to_add to start addition
vbase = dm_dict_en[vecs_to_add[0]]
for vec in vecs_to_add[1:]:
vbase = vbase + dm_dict_en[vec]
vbase = normalise(vbase)
freqs = compute_freq_vector(words_for_freqs)
return vbase, freqs
def pod_from_scratch(name,url,language,description):
if not db.session.query(Pods).filter_by(url=url).all():
p = Pods(url=url)
db.session.add(p)
db.session.commit()
p = Pods.query.filter(Pods.url == url).first()
p.name = name
p.description = description
p.language = language
#Using compute_query_vector as hack to get vectors from pod's name
vector, freqs = compute_query_vectors(name.lower()+' '+description.lower(), dm_dict_en)
p.DS_vector = convert_to_string(normalise(vector))
word_vector = ""
c = 0
for w in sorted(freqs, key=freqs.get, reverse=True):
word_vector += w + ':' + str(freqs[w]) + ' '
c += 1
if c == 300:
break
p.word_vector = word_vector
if not p.registered:
p.registered = False
db.session.commit()