-
Notifications
You must be signed in to change notification settings - Fork 0
/
rel_predictor.py
46 lines (36 loc) · 1.27 KB
/
rel_predictor.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
import numpy as np
from sklearn.metrics import r2_score
from sklearn.model_selection import train_test_split
from keras.layers import Activation, Dense
from keras.models import Sequential, load_model
from web_processor import WebProcessor
class RelPredictor:
def __init__(self, query, processor=None, model=None):
self.query = query
if model:
self.model = model
else:
self.model = load_model('model/default_model.h5')
if not processor:
self.processor = WebProcessor(query=self.query)
else:
self.processor = processor
def train_model(self, label, features):
pass
def save_model(self):
self.model.save('model/default_model.h5')
def get_relevance(self, url):
self.processor.crawl_website(url)
tag_text = self.processor.extract_by_tags()
tf = self.processor.get_tfidf(tag_text)
return self.model.predict(np.array([
tf,
]))[0][0]
def main():
query = 'artificial intelligence'.split()
rp = RelPredictor(query=query)
score = rp.get_relevance(
'https://en.wikipedia.org/wiki/Artificial_intelligence')
print(score)
if __name__ == '__main__':
main()