def isDifferentModel_2(pair, feature):
if feature in products[pair[0]] and feature in products[pair[1]]:
return simfunctions.overlap_coefficient(
tokenizers.qgram(delNonNumber(pair[0], feature), 3),
tokenizers.qgram(delNonNumber(pair[1], feature), 3))
else:
return noneValue
def setMatchNumKeyField(pair, fields):
data1 = nonNum_re.sub(' ', str(products[pair[1]])).strip().split()
if data1:
for d1 in data1:
if len(d1) >= MIN_KEY_LEN:
if PARTIAL_KEY:
data2 = tokenizers.qgram(d1, MIN_KEY_LEN)
else:
data2 = tokenizers.qgram(d1, len(d1))
for d2 in data2:
for field in fields:
try:
if (d2 in products[pair[0]][field][0]):
return 1
except:
continue
return 0
def test_qgrams_valid(self):
self.assertEqual(qgram(''), [])
self.assertEqual(qgram('a'), [])
self.assertEqual(qgram('aa'), ['aa'])
self.assertEqual(qgram('database'), ['da', 'at', 'ta', 'ab', 'ba', 'as', 'se'])
self.assertEqual(qgram('d', 1), ['d'])
self.assertEqual(qgram('database', 3), ['dat', 'ata', 'tab', 'aba', 'bas', 'ase'])
def test_qgrams_valid(self):
self.assertEqual(qgram(''), [])
self.assertEqual(qgram('a'), [])
self.assertEqual(qgram('aa'), ['aa'])
self.assertEqual(qgram('database'),
['da', 'at', 'ta', 'ab', 'ba', 'as', 'se'])
self.assertEqual(qgram('d', 1), ['d'])
self.assertEqual(qgram('database', 3),
['dat', 'ata', 'tab', 'aba', 'bas', 'ase'])
def generate_feature(file_name):
lines = stage3_helper.read_file(file_name)
features = []
labels = []
all_names = []
for line in lines:
json1, json2, label = stage3_helper.read_jsons_label_from_line(line)
string1, string2 = stage3_helper.get_attribute_from_jsons(
json1, json2, product_name)
all_names.append(tokenizers.whitespace(string1))
all_names.append(tokenizers.whitespace(string2))
for line in lines:
json1, json2, label = stage3_helper.read_jsons_label_from_line(line)
feature = []
# TODO: Add more features and optimize features.
# product_type
string1, string2 = stage3_helper.get_attribute_from_jsons(
json1, json2, product_type)
string1 = string1.lower()
string2 = string2.lower()
feature.append(
simfunctions.jaccard(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.jaro_winkler(string1, string2, prefix_weight=0.1))
feature.append(
simfunctions.jaro(
tokenizers.whitespace(string1)[0],
tokenizers.whitespace(string2)[0]))
# if len(string1) == len(string2):
# feature.append(simfunctions.hamming_distance(string1, string2))
# else:
# feature.append(5)
feature.append(
simfunctions.cosine(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.overlap_coefficient(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.monge_elkan(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.tfidf(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(len(string1))
feature.append(len(string2))
feature.append(len(string1) - len(string2))
feature.append(len(tokenizers.whitespace(string1)))
feature.append(len(tokenizers.whitespace(string2)))
# product_name
string1, string2 = stage3_helper.get_attribute_from_jsons(
json1, json2, product_name)
string1 = string1.lower()
string2 = string2.lower()
feature.append(
simfunctions.jaccard(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.jaro_winkler(string1, string2, prefix_weight=0.1))
if len(string1) == len(string2):
feature.append(simfunctions.hamming_distance(string1, string2))
else:
feature.append(5)
feature.append(
simfunctions.cosine(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.overlap_coefficient(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.monge_elkan(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.tfidf(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(len(string1))
feature.append(len(string2))
feature.append(len(string1) - len(string2))
feature.append(
simfunctions.jaro(
tokenizers.whitespace(string1)[0],
tokenizers.whitespace(string2)[0]))
feature.append(len(tokenizers.whitespace(string1)))
feature.append(len(tokenizers.whitespace(string2)))
# product_segment
string1, string2 = stage3_helper.get_attribute_from_jsons(
json1, json2, product_segment)
string1 = string1.lower()
string2 = string2.lower()
feature.append(
simfunctions.jaccard(tokenizers.qgram(string1, 3),
tokenizers.qgram(string2, 3)))
feature.append(
simfunctions.jaro_winkler(string1, string2, prefix_weight=0.1))
# if len(string1) == len(string2):
# feature.append(simfunctions.hamming_distance(string1, string2))
# else:
# feature.append(5)
feature.append(
simfunctions.cosine(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.overlap_coefficient(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.monge_elkan(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.tfidf(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.jaro(
tokenizers.whitespace(string1)[0],
tokenizers.whitespace(string2)[0]))
feature.append(len(string1))
feature.append(len(string2))
feature.append(len(string1) - len(string2))
feature.append(len(tokenizers.whitespace(string1)))
feature.append(len(tokenizers.whitespace(string2)))
# product_long_description
string1, string2 = stage3_helper.get_attribute_from_jsons(
json1, json2, product_long_description)
if string1 is None or string2 is None:
feature.append(0.5)
feature.append(0)
feature.append(0)
feature.append(0)
feature.append(0)
# feature.append(0)
# feature.append(0)
# feature.append(0)
# feature.append(0)
else:
string1 = string1.lower()
string2 = string2.lower()
string1 = stage3_helper.cleanhtml(string1)
string2 = stage3_helper.cleanhtml(string2)
string1 = stage3_helper.clean_stop_word(string1)
string2 = stage3_helper.clean_stop_word(string2)
feature.append(
simfunctions.jaccard(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
# feature.append(simfunctions.jaro_winkler(string1, string2, prefix_weight=0.1))
feature.append(
simfunctions.overlap_coefficient(
tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
# feature.append(simfunctions.monge_elkan(tokenizers.whitespace(string1), tokenizers.whitespace(string2)))
# feature.append(simfunctions.tfidf(tokenizers.whitespace(string1), tokenizers.whitespace(string2)))
feature.append(len(string1))
feature.append(len(string2))
feature.append(len(string1) - len(string2))
# product_brand
string1, string2 = stage3_helper.get_attribute_from_jsons(
json1, json2, product_brand)
string1_name, string2_name = stage3_helper.get_attribute_from_jsons(
json1, json2, product_name)
if string1 is None or string1 == '':
string1 = get_predict_brand(string1_name)
if string2 is None or string2 == '':
string2 = get_predict_brand(string2_name)
if string1 is None or string2 is None:
feature.append(0)
feature.append(0)
feature.append(0)
feature.append(0)
feature.append(0)
feature.append(0)
feature.append(0)
feature.append(0)
else:
feature.append(
simfunctions.jaccard(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.jaro_winkler(string1, string2, prefix_weight=0.1))
feature.append(
simfunctions.overlap_coefficient(
tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.monge_elkan(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(
simfunctions.tfidf(tokenizers.whitespace(string1),
tokenizers.whitespace(string2)))
feature.append(len(string1))
feature.append(len(string2))
feature.append(len(string1) - len(string2))
#feature.append(simfunctions.jaro(tokenizers.whitespace(string1)[0], tokenizers.whitespace(string2)[0]))
# Contains similar model names.
string1, string2 = stage3_helper.get_attribute_from_jsons(
json1, json2, product_name)
string1 = string1.lower()
string2 = string2.lower()
model_strs1 = stage3_helper.find_model_str(string1)
model_strs2 = stage3_helper.find_model_str(string2)
# share_model_str = False
# for model in model_strs1:
# if model.lower() in string2.lower():
# share_model_str = True
# for model in model_strs2:
# if model.lower() in string1.lower():
# share_model_str = True
# if share_model_str:
# feature.append(1)
# else:
# feature.append(0)
if len(model_strs1) > 0 and len(model_strs2) > 0:
feature.append(simfunctions.jaccard(model_strs1, model_strs2))
else:
feature.append(0.5)
feature.append(len(model_strs1))
feature.append(len(model_strs2))
feature.append(len(model_strs1) - len(string2))
# Other features.
common = 0
common_score = 0.0
for item in json1:
if item in json2:
common += 1
common_score += simfunctions.jaccard(
tokenizers.whitespace(json1[item][0]),
tokenizers.whitespace(json2[item][0]))
common_score = common_score / common
feature.append(len(json1))
feature.append(len(json2))
feature.append(len(json1) - len(json2))
feature.append(common)
feature.append(common_score)
feature.append(len(json.dumps(json1)))
feature.append(len(json.dumps(json2)))
feature.append(len(json.dumps(json1)) - len(json.dumps(json2)))
feature.append(
simfunctions.jaccard(tokenizers.whitespace(json.dumps(json1)),
tokenizers.whitespace(json.dumps(json2))))
# Add one feature and label.
features.append(feature)
labels.append(stage3_helper.get_01_from_label(label))
return features, labels, lines
def generate_feature(filename):
productName_courpus = []
brand_courpus = []
with open(filename, 'r') as f:
for line in f:
list_line = line.split('?')
attribute_id1 = json.loads(list_line[2], encoding='latin-1')
attribute_id2 = json.loads(list_line[4], encoding='latin-1')
if "Product Name" in attribute_id1:
productName_courpus.append(
tokenizers.delimiter(attribute_id1["Product Name"][0]))
if "Product Name" in attribute_id2:
productName_courpus.append(
tokenizers.delimiter(attribute_id2["Product Name"][0]))
if "Brand" in attribute_id1:
brand_courpus.append(
tokenizers.delimiter(attribute_id1["Brand"][0]))
if "Brand" in attribute_id2:
brand_courpus.append(
tokenizers.delimiter(attribute_id2["Brand"][0]))
feature_matrix = []
with open(filename, 'r') as f:
i = 1
for line in f:
list_line = line.split('?')
attribute_id1 = json.loads(list_line[2], encoding='latin-1')
attribute_id2 = json.loads(list_line[4], encoding='latin-1')
print 'Generate features for pair', i
i = i + 1
instance = []
#Product Name 4
if ("Product Name" in attribute_id1
and "Product Name" in attribute_id2):
jaccard_productName = simfunctions.jaccard(
tokenizers.delimiter(attribute_id1["Product Name"][0]),
tokenizers.delimiter(attribute_id2["Product Name"][0]))
jaccard3gram_productName = simfunctions.jaccard(
tokenizers.qgram(attribute_id1["Product Name"][0], 3),
tokenizers.qgram(attribute_id2["Product Name"][0], 3))
tfidf_productName = simfunctions.tfidf(
tokenizers.delimiter(attribute_id1["Product Name"][0]),
tokenizers.delimiter(attribute_id2["Product Name"][0]),
productName_courpus)
edit_productName = simfunctions.levenshtein(
attribute_id1["Product Name"][0],
attribute_id2["Product Name"][0])
edit_productName = 1 - edit_productName / max(
len(attribute_id1["Product Name"][0]),
len(attribute_id2["Product Name"][0]))
else:
jaccard_productName = 0
jaccard3gram_productName = 0
tfidf_productName = 0
edit_productName = 0
instance += [
jaccard_productName, jaccard3gram_productName,
tfidf_productName, edit_productName
]
#Manufacturer 3
if ("Manufacturer" in attribute_id1
and "Manufacturer" in attribute_id2):
jaccard_manufacturer = simfunctions.jaccard(
tokenizers.delimiter(attribute_id1["Manufacturer"][0]),
tokenizers.delimiter(attribute_id2["Manufacturer"][0]))
jaccard3gram_manufacturer = simfunctions.jaccard(
tokenizers.qgram(attribute_id1["Manufacturer"][0], 3),
tokenizers.qgram(attribute_id2["Manufacturer"][0], 3))
tfidf_manufacturer = simfunctions.tfidf(
tokenizers.delimiter(attribute_id1["Manufacturer"][0]),
tokenizers.delimiter(attribute_id2["Manufacturer"][0]))
else:
jaccard_manufacturer = 0
jaccard3gram_manufacturer = 0
tfidf_manufacturer = 0
instance += [
jaccard_manufacturer, jaccard3gram_manufacturer,
tfidf_manufacturer
]
#Color 3
if ("Color" in attribute_id1 and "Color" in attribute_id2):
jaccard_color = simfunctions.jaccard(
tokenizers.delimiter(attribute_id1["Color"][0]),
tokenizers.delimiter(attribute_id2["Color"][0]))
jaccard3gram_color = simfunctions.jaccard(
tokenizers.qgram(attribute_id1["Color"][0], 3),
tokenizers.qgram(attribute_id2["Color"][0], 3))
tfidf_color = simfunctions.tfidf(
tokenizers.delimiter(attribute_id1["Color"][0]),
tokenizers.delimiter(attribute_id2["Color"][0]))
else:
jaccard_color = 0
jaccard3gram_color = 0
tfidf_color = 0
instance += [jaccard_color, jaccard3gram_color, tfidf_color]
#Product Type 3
if ("Product Type" in attribute_id1
and "Product Type" in attribute_id2):
jaccard_productType = simfunctions.jaccard(
tokenizers.delimiter(attribute_id1["Product Type"][0]),
tokenizers.delimiter(attribute_id2["Product Type"][0]))
jaccard3gram_productType = simfunctions.jaccard(
tokenizers.qgram(attribute_id1["Product Type"][0], 3),
tokenizers.qgram(attribute_id2["Product Type"][0], 3))
tfidf_productType = simfunctions.tfidf(
tokenizers.delimiter(attribute_id1["Product Type"][0]),
tokenizers.delimiter(attribute_id2["Product Type"][0]))
else:
jaccard_productType = 0
jaccard3gram_productType = 0
tfidf_productType = 0
instance += [
jaccard_productType, jaccard3gram_productType,
tfidf_productType
]
#Product Segment 3
if "Product Segment" in attribute_id1 and "Product Segment" in attribute_id2:
jaccard_productSegment = simfunctions.jaccard(
tokenizers.delimiter(attribute_id1["Product Segment"][0]),
tokenizers.delimiter(attribute_id2["Product Segment"][0]))
jaccard3gram_productSegment = simfunctions.jaccard(
tokenizers.qgram(attribute_id1["Product Segment"][0], 3),
tokenizers.qgram(attribute_id2["Product Segment"][0], 3))
if (attribute_id1["Product Segment"][0] ==
attribute_id2["Product Segment"][0]):
exactMatch_productSegment = 1
else:
exactMatch_productSegment = 0
else:
exactMatch_productSegment = 0
jaccard_productSegment = 0
jaccard3gram_productSegment = 0
instance += [
exactMatch_productSegment, jaccard_productSegment,
jaccard3gram_productSegment
]
#Brand 4
if ("Brand" in attribute_id1 and "Brand" in attribute_id2):
jaccard_brand = simfunctions.jaccard(
tokenizers.delimiter(attribute_id1["Brand"][0]),
tokenizers.delimiter(attribute_id2["Brand"][0]))
jaccard3gram_brand = simfunctions.jaccard(
tokenizers.qgram(attribute_id1["Brand"][0], 3),
tokenizers.qgram(attribute_id2["Brand"][0], 3))
edit_brand = simfunctions.levenshtein(
attribute_id1["Brand"][0], attribute_id2["Brand"][0])
edit_brand = 1 - edit_brand / max(
len(attribute_id1["Brand"][0]),
len(attribute_id2["Brand"][0]))
tfidf_brand = simfunctions.tfidf(
tokenizers.delimiter(attribute_id1["Brand"][0]),
tokenizers.delimiter(attribute_id2["Brand"][0]),
brand_courpus)
else:
jaccard3gram_brand = 0
jaccard_brand = 0
edit_brand = 0
tfidf_brand = 0
instance += [
jaccard_brand, jaccard3gram_brand, edit_brand, tfidf_brand
]
#Category 2
if ("Category" in attribute_id1 and "Category" in attribute_id2):
jaccard_category = simfunctions.jaccard(
tokenizers.delimiter(attribute_id1["Category"][0]),
tokenizers.delimiter(attribute_id2["Category"][0]))
jaccard3gram_category = simfunctions.jaccard(
tokenizers.qgram(attribute_id1["Category"][0], 3),
tokenizers.qgram(attribute_id2["Category"][0], 3))
else:
jaccard_category = 0
jaccard3gram_category = 0
instance += [jaccard_category, jaccard3gram_category]
#Long Description 3
if ("Product Long Description" in attribute_id1
and "Product Long Description" in attribute_id2):
tfidf_long_description = simfunctions.tfidf(
tokenizers.delimiter(
attribute_id1["Product Long Description"][0]),
tokenizers.delimiter(
attribute_id2["Product Long Description"][0]))
jaccard_long_description = simfunctions.jaccard(
tokenizers.delimiter(
attribute_id1["Product Long Description"][0]),
tokenizers.delimiter(
attribute_id2["Product Long Description"][0]))
jaccard3_long_description = simfunctions.jaccard(
tokenizers.qgram(
attribute_id1["Product Long Description"][0], 3),
tokenizers.qgram(
attribute_id2["Product Long Description"][0], 3))
else:
tfidf_long_description = 0
jaccard_long_description = 0
jaccard3_long_description = 0
instance += [
tfidf_long_description, jaccard_long_description,
jaccard3_long_description
]
#Short Description 3
if ("Product Short Description" in attribute_id1
and "Product Short Description" in attribute_id2):
jaccard_short_description = simfunctions.jaccard(
tokenizers.delimiter(
attribute_id1["Product Short Description"][0]),
tokenizers.delimiter(
attribute_id2["Product Short Description"][0]))
jaccard3_short_description = simfunctions.jaccard(
tokenizers.qgram(
attribute_id1["Product Short Description"][0], 3),
tokenizers.qgram(
attribute_id2["Product Short Description"][0], 3))
tfidf_short_description = simfunctions.tfidf(
tokenizers.delimiter(
attribute_id1["Product Short Description"][0]),
tokenizers.delimiter(
attribute_id2["Product Short Description"][0]))
else:
jaccard_short_description = 0
jaccard3_short_description = 0
tfidf_short_description = 0
instance += [
jaccard_short_description, jaccard3_short_description,
tfidf_short_description
]
#Other in long 8
if ("Product Name" in attribute_id1
and "Product Long Description" in attribute_id2):
name_set = tokenizers.delimiter(
attribute_id1["Product Name"][0])
des = attribute_id2["Product Long Description"][0]
count = 0
for name in name_set:
if name in des:
count = count + 1
name1_in_des2 = count / len(name_set)
else:
name1_in_des2 = 0
instance += [name1_in_des2]
if ("Product Long Description" in attribute_id1
and "Product Name" in attribute_id2):
name_set = tokenizers.delimiter(
attribute_id2["Product Name"][0])
des = attribute_id1["Product Long Description"][0]
count = 0
for name in name_set:
if name in des:
count = count + 1
name2_in_des1 = count / len(name_set)
else:
name2_in_des1 = 0
instance += [name2_in_des1]
if ("Brand" in attribute_id1
and "Product Long Description" in attribute_id2):
brand_set = tokenizers.delimiter(attribute_id1["Brand"][0])
des = attribute_id2["Product Long Description"][0]
count = 0
for brand in brand_set:
if brand in des:
count = count + 1
brand1_in_des2 = count / len(brand_set)
else:
brand1_in_des2 = 0
instance += [brand1_in_des2]
if ("Brand" in attribute_id2
and "Product Long Description" in attribute_id1):
brand_set = tokenizers.delimiter(attribute_id2["Brand"][0])
des = attribute_id1["Product Long Description"][0]
count = 0
for brand in brand_set:
if brand in des:
count = count + 1
brand2_in_des1 = count / len(brand_set)
else:
brand2_in_des1 = 0
instance += [brand2_in_des1]
if ("Manufacturer" in attribute_id1
and "Product Long Description" in attribute_id2):
manufacturer_set = tokenizers.delimiter(
attribute_id1["Manufacturer"][0])
des = attribute_id2["Product Long Description"][0]
count = 0
for manufacturer in manufacturer_set:
if manufacturer in des:
count = count + 1
manufacturer1_in_des2 = count / len(manufacturer_set)
else:
manufacturer1_in_des2 = 0
instance += [manufacturer1_in_des2]
if ("Manufacturer" in attribute_id2
and "Product Long Description" in attribute_id1):
manufacturer_set = tokenizers.delimiter(
attribute_id2["Manufacturer"][0])
des = attribute_id1["Product Long Description"][0]
count = 0
for manufacturer in manufacturer_set:
if manufacturer in des:
count = count + 1
manufacturer2_in_des1 = count / len(manufacturer_set)
else:
manufacturer2_in_des1 = 0
instance += [manufacturer2_in_des1]
if ("Product Short Description" in attribute_id1
and "Product Long Description" in attribute_id2):
short_des_set = tokenizers.delimiter(
attribute_id1["Product Short Description"][0])
des = attribute_id2["Product Long Description"][0]
count = 0
for short in short_des_set:
if short in des:
count = count + 1
short1_in_des2 = count / len(short_des_set)
else:
short1_in_des2 = 0
instance += [short1_in_des2]
if ("Product Short Description" in attribute_id2
and "Product Long Description" in attribute_id1):
short_des_set = tokenizers.delimiter(
attribute_id2["Product Short Description"][0])
des = attribute_id1["Product Long Description"][0]
count = 0
for short in short_des_set:
if short in des:
count = count + 1
short2_in_des1 = count / len(short_des_set)
else:
short2_in_des1 = 0
instance += [short2_in_des1]
#Other in short 6
if ("Product Name" in attribute_id1
and "Product Short Description" in attribute_id2):
name_set = tokenizers.delimiter(
attribute_id1["Product Name"][0])
des = attribute_id2["Product Short Description"][0]
count = 0
for name in name_set:
if name in des:
count = count + 1
name1_in_short2 = count / len(name_set)
else:
name1_in_short2 = 0
instance += [name1_in_short2]
if ("Product Short Description" in attribute_id1
and "Product Name" in attribute_id2):
name_set = tokenizers.delimiter(
attribute_id2["Product Name"][0])
des = attribute_id1["Product Short Description"][0]
count = 0
for name in name_set:
if name in des:
count = count + 1
name2_in_short1 = count / len(name_set)
else:
name2_in_short1 = 0
instance += [name2_in_short1]
if ("Brand" in attribute_id1
and "Product Short Description" in attribute_id2):
brand_set = tokenizers.delimiter(attribute_id1["Brand"][0])
des = attribute_id2["Product Short Description"][0]
count = 0
for brand in brand_set:
if brand in des:
count = count + 1
brand1_in_short2 = count / len(brand_set)
else:
brand1_in_short2 = 0
instance += [brand1_in_short2]
if ("Brand" in attribute_id2
and "Product Short Description" in attribute_id1):
brand_set = tokenizers.delimiter(attribute_id2["Brand"][0])
des = attribute_id1["Product Short Description"][0]
count = 0
for brand in brand_set:
if brand in des:
count = count + 1
brand2_in_short1 = count / len(brand_set)
else:
brand2_in_short1 = 0
instance += [brand2_in_short1]
if ("Manufacturer" in attribute_id1
and "Product Short Description" in attribute_id2):
manufacturer_set = tokenizers.delimiter(
attribute_id1["Manufacturer"][0])
des = attribute_id2["Product Short Description"][0]
count = 0
for manufacturer in manufacturer_set:
if manufacturer in des:
count = count + 1
manufacturer1_in_short2 = count / len(manufacturer_set)
else:
manufacturer1_in_short2 = 0
instance += [manufacturer1_in_short2]
if ("Manufacturer" in attribute_id2
and "Product Short Description" in attribute_id1):
manufacturer_set = tokenizers.delimiter(
attribute_id2["Manufacturer"][0])
des = attribute_id1["Product Short Description"][0]
count = 0
for manufacturer in manufacturer_set:
if manufacturer in des:
count = count + 1
manufacturer2_in_short1 = count / len(manufacturer_set)
else:
manufacturer2_in_short1 = 0
instance += [manufacturer2_in_short1]
#new 15
if ("Manufacturer Part Number" in attribute_id1
and "Product Long Description" in attribute_id2):
manu_part_number_set = tokenizers.delimiter(
attribute_id1["Manufacturer Part Number"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for manu_part in manu_part_number_set:
if manu_part in des_set:
count = count + 1
manu_part1_in_des2 = count / len(manu_part_number_set)
else:
manu_part1_in_des2 = 0
instance += [manu_part1_in_des2]
if ("Manufacturer Part Number" in attribute_id2
and "Product Long Description" in attribute_id1):
manu_part_number_set = tokenizers.delimiter(
attribute_id2["Manufacturer Part Number"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for manu_part in manu_part_number_set:
if manu_part in des_set:
count = count + 1
manu_part2_in_des1 = count / len(manu_part_number_set)
else:
manu_part2_in_des1 = 0
instance += [manu_part2_in_des1]
if ("Assembled Product Length" in attribute_id1
and "Product Long Description" in attribute_id2):
length_set = tokenizers.delimiter(
attribute_id1["Assembled Product Length"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for length in length_set:
if length in des_set:
count = count + 1
length1_in_des2 = count / len(length_set)
else:
length1_in_des2 = 0
instance += [length1_in_des2]
if ("Assembled Product Length" in attribute_id2
and "Product Long Description" in attribute_id1):
length_set = tokenizers.delimiter(
attribute_id2["Assembled Product Length"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for length in length_set:
if length in des_set:
count = count + 1
length2_in_des1 = count / len(length_set)
else:
length2_in_des1 = 0
instance += [length2_in_des1]
if ("Assembled Product Width" in attribute_id1
and "Product Long Description" in attribute_id2):
width_set = tokenizers.delimiter(
attribute_id1["Assembled Product Width"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for width in width_set:
if width in des_set:
count = count + 1
width1_in_des2 = count / len(width_set)
else:
width1_in_des2 = 0
instance += [width1_in_des2]
if ("Assembled Product Width" in attribute_id2
and "Product Long Description" in attribute_id1):
width_set = tokenizers.delimiter(
attribute_id2["Assembled Product Width"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for width in width_set:
if width in des_set:
count = count + 1
width2_in_des1 = count / len(width_set)
else:
width2_in_des1 = 0
instance += [width2_in_des1]
if ("Assembled Product Height" in attribute_id1
and "Product Long Description" in attribute_id2):
height_set = tokenizers.delimiter(
attribute_id1["Assembled Product Height"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for height in height_set:
if height in des_set:
count = count + 1
height1_in_des2 = count / len(height_set)
else:
height1_in_des2 = 0
instance += [height1_in_des2]
if ("Assembled Product Height" in attribute_id2
and "Product Long Description" in attribute_id1):
height_set = tokenizers.delimiter(
attribute_id2["Assembled Product Height"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for height in height_set:
if height in des_set:
count = count + 1
height2_in_des1 = count / len(height_set)
else:
height2_in_des1 = 0
instance += [height2_in_des1]
if ("Type" in attribute_id1
and "Product Long Description" in attribute_id2):
type_set = tokenizers.delimiter(attribute_id1["Type"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for type in type_set:
if type in des_set:
count = count + 1
type1_in_des2 = count / len(type_set)
else:
type1_in_des2 = 0
instance += [type1_in_des2]
if ("Type" in attribute_id2
and "Product Long Description" in attribute_id1):
type_set = tokenizers.delimiter(attribute_id2["Type"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for type in type_set:
if type in des_set:
count = count + 1
type2_in_des1 = count / len(type_set)
else:
type2_in_des1 = 0
instance += [type2_in_des1]
if ("Operating System" in attribute_id1
and "Product Long Description" in attribute_id2):
op_set = tokenizers.delimiter(
attribute_id1["Operating System"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for op in op_set:
if op in op_set:
count = count + 1
op1_in_des2 = count / len(op_set)
else:
op1_in_des2 = 0
instance += [op1_in_des2]
if ("Operating System" in attribute_id2
and "Product Long Description" in attribute_id1):
op_set = tokenizers.delimiter(
attribute_id2["Operating System"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for op in op_set:
if op in op_set:
count = count + 1
op2_in_des1 = count / len(op_set)
else:
op2_in_des1 = 0
instance += [op2_in_des1]
if ("Screen Size" in attribute_id1
and "Product Long Description" in attribute_id2):
ss_set = tokenizers.delimiter(attribute_id1["Screen Size"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for ss in ss_set:
if ss in ss_set:
count = count + 1
ss1_in_des2 = count / len(ss_set)
else:
ss1_in_des2 = 0
instance += [ss1_in_des2]
if ("Screen Size" in attribute_id2
and "Product Long Description" in attribute_id1):
ss_set = tokenizers.delimiter(attribute_id2["Screen Size"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for ss in ss_set:
if ss in ss_set:
count = count + 1
ss2_in_des1 = count / len(ss_set)
else:
ss2_in_des1 = 0
instance += [ss2_in_des1]
if "Product Long Description" in attribute_id2:
all_set = []
for key in attribute_id1:
if key is not "Product Long Description":
value_list = tokenizers.delimiter(
attribute_id1[key][0])
for v in value_list:
all_set.append(v)
des = attribute_id2["Product Long Description"][0]
count = 0
for a in all_set:
if a in des:
count += 1
all1_in_des2 = count / len(all_set)
else:
all1_in_des2 = 0
instance += [all1_in_des2]
feature_matrix.append(instance)
return feature_matrix
id2 = pair[1]
attribute_id1 = product_dict[id1]
attribute_id2 = product_dict[id2]
id.append([id1,id2])
# class label
if (match_dict[pair] == 'MATCH'):
classlabels.append(1)
else:
classlabels.append(0)
####feature: Product Name ---- Jaccard Score (word boudary, 3-gram), edit distance, tf/idf
if ("Product Name" in attribute_id1 and "Product Name" in attribute_id2):
jaccard_productName = simfunctions.jaccard(tokenizers.delimiter(attribute_id1["Product Name"][0]), tokenizers.delimiter(attribute_id2["Product Name"][0]))
jaccard3gram_productName = simfunctions.jaccard(tokenizers.qgram(attribute_id1["Product Name"][0], 3), tokenizers.qgram(attribute_id2["Product Name"][0], 3))
tfidf_productName = simfunctions.tfidf(tokenizers.delimiter(attribute_id1["Product Name"][0]), tokenizers.delimiter(attribute_id2["Product Name"][0]), productName_courpus)
edit_productName = simfunctions.levenshtein(attribute_id1["Product Name"][0], attribute_id2["Product Name"][0])
edit_productName = 1 - edit_productName/max(len(attribute_id1["Product Name"][0]), len(attribute_id2["Product Name"][0]))
else:
jaccard_productName = 0
jaccard3gram_productName = 0
tfidf_productName = 0
edit_productName = 0
####feature: Manufacturer
if ("Manufacturer" in attribute_id1 and "Manufacturer" in attribute_id2):
jaccard_manufacturer = simfunctions.jaccard(tokenizers.delimiter(attribute_id1["Manufacturer"][0]), tokenizers.delimiter(attribute_id2["Manufacturer"][0]))
jaccard3gram_manufacturer = simfunctions.jaccard(tokenizers.qgram(attribute_id1["Manufacturer"][0], 3), tokenizers.qgram(attribute_id2["Manufacturer"][0], 3))
tfidf_manufacturer = simfunctions.tfidf(tokenizers.delimiter(attribute_id1["Manufacturer"][0]), tokenizers.delimiter(attribute_id2["Manufacturer"][0]))
else:
def generate_feature(filename):
productName_courpus = []
brand_courpus = []
with open(filename, 'r') as f:
for line in f:
list_line = line.split('?')
attribute_id1 = json.loads(list_line[2], encoding = 'latin-1')
attribute_id2 = json.loads(list_line[4], encoding = 'latin-1')
if "Product Name" in attribute_id1:
productName_courpus.append(tokenizers.delimiter(attribute_id1["Product Name"][0]))
if "Product Name" in attribute_id2:
productName_courpus.append(tokenizers.delimiter(attribute_id2["Product Name"][0]))
if "Brand" in attribute_id1:
brand_courpus.append(tokenizers.delimiter(attribute_id1["Brand"][0]))
if "Brand" in attribute_id2:
brand_courpus.append(tokenizers.delimiter(attribute_id2["Brand"][0]))
feature_matrix = []
with open(filename, 'r') as f:
i = 1
for line in f:
list_line = line.split('?')
attribute_id1 = json.loads(list_line[2], encoding = 'latin-1')
attribute_id2 = json.loads(list_line[4], encoding = 'latin-1')
print 'Generate features for pair', i
i = i+1
instance = []
#Product Name 4
if ("Product Name" in attribute_id1 and "Product Name" in attribute_id2):
jaccard_productName = simfunctions.jaccard(tokenizers.delimiter(attribute_id1["Product Name"][0]), tokenizers.delimiter(attribute_id2["Product Name"][0]))
jaccard3gram_productName = simfunctions.jaccard(tokenizers.qgram(attribute_id1["Product Name"][0], 3), tokenizers.qgram(attribute_id2["Product Name"][0], 3))
tfidf_productName = simfunctions.tfidf(tokenizers.delimiter(attribute_id1["Product Name"][0]), tokenizers.delimiter(attribute_id2["Product Name"][0]), productName_courpus)
edit_productName = simfunctions.levenshtein(attribute_id1["Product Name"][0], attribute_id2["Product Name"][0])
edit_productName = 1 - edit_productName/max(len(attribute_id1["Product Name"][0]), len(attribute_id2["Product Name"][0]))
else:
jaccard_productName = 0
jaccard3gram_productName = 0
tfidf_productName = 0
edit_productName = 0
instance += [jaccard_productName, jaccard3gram_productName, tfidf_productName, edit_productName]
#Manufacturer 3
if ("Manufacturer" in attribute_id1 and "Manufacturer" in attribute_id2):
jaccard_manufacturer = simfunctions.jaccard(tokenizers.delimiter(attribute_id1["Manufacturer"][0]), tokenizers.delimiter(attribute_id2["Manufacturer"][0]))
jaccard3gram_manufacturer = simfunctions.jaccard(tokenizers.qgram(attribute_id1["Manufacturer"][0], 3), tokenizers.qgram(attribute_id2["Manufacturer"][0], 3))
tfidf_manufacturer = simfunctions.tfidf(tokenizers.delimiter(attribute_id1["Manufacturer"][0]), tokenizers.delimiter(attribute_id2["Manufacturer"][0]))
else:
jaccard_manufacturer = 0
jaccard3gram_manufacturer = 0
tfidf_manufacturer = 0
instance += [jaccard_manufacturer, jaccard3gram_manufacturer, tfidf_manufacturer]
#Color 3
if ("Color" in attribute_id1 and "Color" in attribute_id2):
jaccard_color = simfunctions.jaccard(tokenizers.delimiter(attribute_id1["Color"][0]), tokenizers.delimiter(attribute_id2["Color"][0]))
jaccard3gram_color = simfunctions.jaccard(tokenizers.qgram(attribute_id1["Color"][0], 3), tokenizers.qgram(attribute_id2["Color"][0], 3))
tfidf_color = simfunctions.tfidf(tokenizers.delimiter(attribute_id1["Color"][0]), tokenizers.delimiter(attribute_id2["Color"][0]))
else:
jaccard_color = 0
jaccard3gram_color = 0
tfidf_color = 0
instance += [jaccard_color, jaccard3gram_color, tfidf_color]
#Product Type 3
if ("Product Type" in attribute_id1 and "Product Type" in attribute_id2):
jaccard_productType = simfunctions.jaccard(tokenizers.delimiter(attribute_id1["Product Type"][0]),tokenizers.delimiter(attribute_id2["Product Type"][0]))
jaccard3gram_productType = simfunctions.jaccard(tokenizers.qgram(attribute_id1["Product Type"][0], 3),tokenizers.qgram(attribute_id2["Product Type"][0], 3))
tfidf_productType = simfunctions.tfidf(tokenizers.delimiter(attribute_id1["Product Type"][0]),tokenizers.delimiter(attribute_id2["Product Type"][0]))
else:
jaccard_productType = 0
jaccard3gram_productType = 0
tfidf_productType = 0
instance += [jaccard_productType, jaccard3gram_productType, tfidf_productType]
#Product Segment 3
if "Product Segment" in attribute_id1 and "Product Segment" in attribute_id2:
jaccard_productSegment = simfunctions.jaccard(tokenizers.delimiter(attribute_id1["Product Segment"][0]),tokenizers.delimiter(attribute_id2["Product Segment"][0]))
jaccard3gram_productSegment= simfunctions.jaccard(tokenizers.qgram(attribute_id1["Product Segment"][0], 3),tokenizers.qgram(attribute_id2["Product Segment"][0], 3))
if (attribute_id1["Product Segment"][0] == attribute_id2["Product Segment"][0]):
exactMatch_productSegment = 1
else:
exactMatch_productSegment = 0
else:
exactMatch_productSegment = 0
jaccard_productSegment = 0
jaccard3gram_productSegment = 0
instance += [exactMatch_productSegment, jaccard_productSegment, jaccard3gram_productSegment]
#Brand 4
if ("Brand" in attribute_id1 and "Brand" in attribute_id2):
jaccard_brand = simfunctions.jaccard(tokenizers.delimiter(attribute_id1["Brand"][0]),tokenizers.delimiter(attribute_id2["Brand"][0]))
jaccard3gram_brand = simfunctions.jaccard(tokenizers.qgram(attribute_id1["Brand"][0], 3),
tokenizers.qgram(attribute_id2["Brand"][0], 3))
edit_brand = simfunctions.levenshtein(attribute_id1["Brand"][0], attribute_id2["Brand"][0])
edit_brand = 1 - edit_brand/max(len(attribute_id1["Brand"][0]), len(attribute_id2["Brand"][0]))
tfidf_brand = simfunctions.tfidf(tokenizers.delimiter(attribute_id1["Brand"][0]), tokenizers.delimiter(attribute_id2["Brand"][0]), brand_courpus)
else:
jaccard3gram_brand = 0
jaccard_brand = 0
edit_brand = 0
tfidf_brand = 0
instance += [jaccard_brand, jaccard3gram_brand, edit_brand, tfidf_brand]
#Category 2
if ("Category" in attribute_id1 and "Category" in attribute_id2):
jaccard_category = simfunctions.jaccard(tokenizers.delimiter(attribute_id1["Category"][0]),
tokenizers.delimiter(attribute_id2["Category"][0]))
jaccard3gram_category = simfunctions.jaccard(tokenizers.qgram(attribute_id1["Category"][0], 3),
tokenizers.qgram(attribute_id2["Category"][0], 3))
else:
jaccard_category = 0
jaccard3gram_category = 0
instance += [jaccard_category, jaccard3gram_category]
#Long Description 3
if ("Product Long Description" in attribute_id1 and "Product Long Description" in attribute_id2):
tfidf_long_description = simfunctions.tfidf(tokenizers.delimiter(attribute_id1["Product Long Description"][0]), tokenizers.delimiter(attribute_id2["Product Long Description"][0]))
jaccard_long_description = simfunctions.jaccard(tokenizers.delimiter(attribute_id1["Product Long Description"][0]), tokenizers.delimiter(attribute_id2["Product Long Description"][0]))
jaccard3_long_description = simfunctions.jaccard(tokenizers.qgram(attribute_id1["Product Long Description"][0], 3), tokenizers.qgram(attribute_id2["Product Long Description"][0], 3))
else:
tfidf_long_description = 0
jaccard_long_description = 0
jaccard3_long_description = 0
instance += [tfidf_long_description, jaccard_long_description, jaccard3_long_description]
#Short Description 3
if ("Product Short Description" in attribute_id1 and "Product Short Description" in attribute_id2):
jaccard_short_description = simfunctions.jaccard(tokenizers.delimiter(attribute_id1["Product Short Description"][0]), tokenizers.delimiter(attribute_id2["Product Short Description"][0]))
jaccard3_short_description = simfunctions.jaccard(tokenizers.qgram(attribute_id1["Product Short Description"][0], 3), tokenizers.qgram(attribute_id2["Product Short Description"][0], 3))
tfidf_short_description = simfunctions.tfidf(tokenizers.delimiter(attribute_id1["Product Short Description"][0]), tokenizers.delimiter(attribute_id2["Product Short Description"][0]))
else:
jaccard_short_description = 0
jaccard3_short_description = 0
tfidf_short_description = 0
instance += [jaccard_short_description, jaccard3_short_description, tfidf_short_description]
#Other in long 8
if ("Product Name" in attribute_id1 and "Product Long Description" in attribute_id2):
name_set = tokenizers.delimiter(attribute_id1["Product Name"][0])
des = attribute_id2["Product Long Description"][0]
count = 0
for name in name_set:
if name in des:
count = count+1
name1_in_des2 = count/len(name_set)
else:
name1_in_des2 = 0
instance += [name1_in_des2]
if ("Product Long Description" in attribute_id1 and "Product Name" in attribute_id2):
name_set = tokenizers.delimiter(attribute_id2["Product Name"][0])
des = attribute_id1["Product Long Description"][0]
count = 0
for name in name_set:
if name in des:
count = count+1
name2_in_des1 = count/len(name_set)
else:
name2_in_des1 = 0
instance += [name2_in_des1]
if ("Brand" in attribute_id1 and "Product Long Description" in attribute_id2):
brand_set = tokenizers.delimiter(attribute_id1["Brand"][0])
des = attribute_id2["Product Long Description"][0]
count = 0
for brand in brand_set:
if brand in des:
count = count+1
brand1_in_des2 = count/len(brand_set)
else:
brand1_in_des2 = 0
instance += [brand1_in_des2]
if ("Brand" in attribute_id2 and "Product Long Description" in attribute_id1):
brand_set = tokenizers.delimiter(attribute_id2["Brand"][0])
des = attribute_id1["Product Long Description"][0]
count = 0
for brand in brand_set:
if brand in des:
count = count+1
brand2_in_des1 = count/len(brand_set)
else:
brand2_in_des1 = 0
instance += [brand2_in_des1]
if ("Manufacturer" in attribute_id1 and "Product Long Description" in attribute_id2):
manufacturer_set = tokenizers.delimiter(attribute_id1["Manufacturer"][0])
des = attribute_id2["Product Long Description"][0]
count = 0
for manufacturer in manufacturer_set:
if manufacturer in des:
count = count+1
manufacturer1_in_des2 = count/len(manufacturer_set)
else:
manufacturer1_in_des2 = 0
instance += [manufacturer1_in_des2]
if ("Manufacturer" in attribute_id2 and "Product Long Description" in attribute_id1):
manufacturer_set = tokenizers.delimiter(attribute_id2["Manufacturer"][0])
des = attribute_id1["Product Long Description"][0]
count = 0
for manufacturer in manufacturer_set:
if manufacturer in des:
count = count+1
manufacturer2_in_des1 = count/len(manufacturer_set)
else:
manufacturer2_in_des1 = 0
instance += [manufacturer2_in_des1]
if ("Product Short Description" in attribute_id1 and "Product Long Description" in attribute_id2):
short_des_set = tokenizers.delimiter(attribute_id1["Product Short Description"][0])
des = attribute_id2["Product Long Description"][0]
count = 0
for short in short_des_set:
if short in des:
count = count+1
short1_in_des2 = count/len(short_des_set)
else:
short1_in_des2 = 0
instance += [short1_in_des2]
if ("Product Short Description" in attribute_id2 and "Product Long Description" in attribute_id1):
short_des_set = tokenizers.delimiter(attribute_id2["Product Short Description"][0])
des = attribute_id1["Product Long Description"][0]
count = 0
for short in short_des_set:
if short in des:
count = count+1
short2_in_des1 = count/len(short_des_set)
else:
short2_in_des1 = 0
instance += [short2_in_des1]
#Other in short 6
if ("Product Name" in attribute_id1 and "Product Short Description" in attribute_id2):
name_set = tokenizers.delimiter(attribute_id1["Product Name"][0])
des = attribute_id2["Product Short Description"][0]
count = 0
for name in name_set:
if name in des:
count = count+1
name1_in_short2 = count/len(name_set)
else:
name1_in_short2 = 0
instance += [name1_in_short2]
if ("Product Short Description" in attribute_id1 and "Product Name" in attribute_id2):
name_set = tokenizers.delimiter(attribute_id2["Product Name"][0])
des = attribute_id1["Product Short Description"][0]
count = 0
for name in name_set:
if name in des:
count = count+1
name2_in_short1 = count/len(name_set)
else:
name2_in_short1 = 0
instance += [name2_in_short1]
if ("Brand" in attribute_id1 and "Product Short Description" in attribute_id2):
brand_set = tokenizers.delimiter(attribute_id1["Brand"][0])
des = attribute_id2["Product Short Description"][0]
count = 0
for brand in brand_set:
if brand in des:
count = count+1
brand1_in_short2 = count/len(brand_set)
else:
brand1_in_short2 = 0
instance += [brand1_in_short2]
if ("Brand" in attribute_id2 and "Product Short Description" in attribute_id1):
brand_set = tokenizers.delimiter(attribute_id2["Brand"][0])
des = attribute_id1["Product Short Description"][0]
count = 0
for brand in brand_set:
if brand in des:
count = count+1
brand2_in_short1 = count/len(brand_set)
else:
brand2_in_short1 = 0
instance += [brand2_in_short1]
if ("Manufacturer" in attribute_id1 and "Product Short Description" in attribute_id2):
manufacturer_set = tokenizers.delimiter(attribute_id1["Manufacturer"][0])
des = attribute_id2["Product Short Description"][0]
count = 0
for manufacturer in manufacturer_set:
if manufacturer in des:
count = count+1
manufacturer1_in_short2 = count/len(manufacturer_set)
else:
manufacturer1_in_short2 = 0
instance += [manufacturer1_in_short2]
if ("Manufacturer" in attribute_id2 and "Product Short Description" in attribute_id1):
manufacturer_set = tokenizers.delimiter(attribute_id2["Manufacturer"][0])
des = attribute_id1["Product Short Description"][0]
count = 0
for manufacturer in manufacturer_set:
if manufacturer in des:
count = count+1
manufacturer2_in_short1 = count/len(manufacturer_set)
else:
manufacturer2_in_short1 = 0
instance += [manufacturer2_in_short1]
#new 15
if ("Manufacturer Part Number" in attribute_id1 and "Product Long Description" in attribute_id2):
manu_part_number_set = tokenizers.delimiter(attribute_id1["Manufacturer Part Number"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for manu_part in manu_part_number_set:
if manu_part in des_set:
count = count+1
manu_part1_in_des2 = count/len(manu_part_number_set)
else:
manu_part1_in_des2 = 0
instance += [manu_part1_in_des2]
if ("Manufacturer Part Number" in attribute_id2 and "Product Long Description" in attribute_id1):
manu_part_number_set = tokenizers.delimiter(attribute_id2["Manufacturer Part Number"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for manu_part in manu_part_number_set:
if manu_part in des_set:
count = count+1
manu_part2_in_des1 = count/len(manu_part_number_set)
else:
manu_part2_in_des1 = 0
instance += [manu_part2_in_des1]
if ("Assembled Product Length" in attribute_id1 and "Product Long Description" in attribute_id2):
length_set = tokenizers.delimiter(attribute_id1["Assembled Product Length"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for length in length_set:
if length in des_set:
count = count+1
length1_in_des2 = count/len(length_set)
else:
length1_in_des2 = 0
instance += [length1_in_des2]
if ("Assembled Product Length" in attribute_id2 and "Product Long Description" in attribute_id1):
length_set = tokenizers.delimiter(attribute_id2["Assembled Product Length"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for length in length_set:
if length in des_set:
count = count+1
length2_in_des1 = count/len(length_set)
else:
length2_in_des1 = 0
instance += [length2_in_des1]
if ("Assembled Product Width" in attribute_id1 and "Product Long Description" in attribute_id2):
width_set = tokenizers.delimiter(attribute_id1["Assembled Product Width"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for width in width_set:
if width in des_set:
count = count+1
width1_in_des2 = count/len(width_set)
else:
width1_in_des2 = 0
instance += [width1_in_des2]
if ("Assembled Product Width" in attribute_id2 and "Product Long Description" in attribute_id1):
width_set = tokenizers.delimiter(attribute_id2["Assembled Product Width"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for width in width_set:
if width in des_set:
count = count+1
width2_in_des1 = count/len(width_set)
else:
width2_in_des1 = 0
instance += [width2_in_des1]
if ("Assembled Product Height" in attribute_id1 and "Product Long Description" in attribute_id2):
height_set = tokenizers.delimiter(attribute_id1["Assembled Product Height"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for height in height_set:
if height in des_set:
count = count+1
height1_in_des2 = count/len(height_set)
else:
height1_in_des2 = 0
instance += [height1_in_des2]
if ("Assembled Product Height" in attribute_id2 and "Product Long Description" in attribute_id1):
height_set = tokenizers.delimiter(attribute_id2["Assembled Product Height"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for height in height_set:
if height in des_set:
count = count+1
height2_in_des1 = count/len(height_set)
else:
height2_in_des1 = 0
instance += [height2_in_des1]
if ("Type" in attribute_id1 and "Product Long Description" in attribute_id2):
type_set = tokenizers.delimiter(attribute_id1["Type"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for type in type_set:
if type in des_set:
count = count+1
type1_in_des2 = count/len(type_set)
else:
type1_in_des2 = 0
instance += [type1_in_des2]
if ("Type" in attribute_id2 and "Product Long Description" in attribute_id1):
type_set = tokenizers.delimiter(attribute_id2["Type"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for type in type_set:
if type in des_set:
count = count+1
type2_in_des1 = count/len(type_set)
else:
type2_in_des1 = 0
instance += [type2_in_des1]
if ("Operating System" in attribute_id1 and "Product Long Description" in attribute_id2):
op_set = tokenizers.delimiter(attribute_id1["Operating System"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for op in op_set:
if op in op_set:
count = count+1
op1_in_des2 = count/len(op_set)
else:
op1_in_des2 = 0
instance += [op1_in_des2]
if ("Operating System" in attribute_id2 and "Product Long Description" in attribute_id1):
op_set = tokenizers.delimiter(attribute_id2["Operating System"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for op in op_set:
if op in op_set:
count = count+1
op2_in_des1 = count/len(op_set)
else:
op2_in_des1 = 0
instance += [op2_in_des1]
if ("Screen Size" in attribute_id1 and "Product Long Description" in attribute_id2):
ss_set = tokenizers.delimiter(attribute_id1["Screen Size"][0])
des_set = attribute_id2["Product Long Description"][0]
count = 0
for ss in ss_set:
if ss in ss_set:
count = count+1
ss1_in_des2 = count/len(ss_set)
else:
ss1_in_des2 = 0
instance += [ss1_in_des2]
if ("Screen Size" in attribute_id2 and "Product Long Description" in attribute_id1):
ss_set = tokenizers.delimiter(attribute_id2["Screen Size"][0])
des_set = attribute_id1["Product Long Description"][0]
count = 0
for ss in ss_set:
if ss in ss_set:
count = count+1
ss2_in_des1 = count/len(ss_set)
else:
ss2_in_des1 = 0
instance += [ss2_in_des1]
if "Product Long Description" in attribute_id2:
all_set = []
for key in attribute_id1:
if key is not "Product Long Description":
value_list = tokenizers.delimiter(attribute_id1[key][0])
for v in value_list:
all_set.append(v)
des = attribute_id2["Product Long Description"][0]
count = 0
for a in all_set:
if a in des:
count += 1
all1_in_des2 = count/len(all_set)
else:
all1_in_des2 = 0
instance += [all1_in_des2]
feature_matrix.append(instance)
return feature_matrix
def test_qgrams_none(self):
self.assertEqual(qgram(None), [])
attribute_id2 = product_dict[id2]
id.append([id1, id2])
# class label
if (match_dict[pair] == 'MATCH'):
classlabels.append(1)
else:
classlabels.append(0)
####feature: Product Name ---- Jaccard Score (word boudary, 3-gram), edit distance, tf/idf
if ("Product Name" in attribute_id1 and "Product Name" in attribute_id2):
jaccard_productName = simfunctions.jaccard(
tokenizers.delimiter(attribute_id1["Product Name"][0]),
tokenizers.delimiter(attribute_id2["Product Name"][0]))
jaccard3gram_productName = simfunctions.jaccard(
tokenizers.qgram(attribute_id1["Product Name"][0], 3),
tokenizers.qgram(attribute_id2["Product Name"][0], 3))
tfidf_productName = simfunctions.tfidf(
tokenizers.delimiter(attribute_id1["Product Name"][0]),
tokenizers.delimiter(attribute_id2["Product Name"][0]),
productName_courpus)
edit_productName = simfunctions.levenshtein(
attribute_id1["Product Name"][0], attribute_id2["Product Name"][0])
edit_productName = 1 - edit_productName / max(
len(attribute_id1["Product Name"][0]),
len(attribute_id2["Product Name"][0]))
else:
jaccard_productName = 0
jaccard3gram_productName = 0
tfidf_productName = 0
edit_productName = 0
def test_qgrams_none(self):
self.assertEqual(qgram(None), [])