forked from azk0019/CourseProject
-
Notifications
You must be signed in to change notification settings - Fork 0
Expand file tree
/
Copy pathmodel.py
More file actions
37 lines (33 loc) · 1.55 KB
/
model.py
File metadata and controls
37 lines (33 loc) · 1.55 KB
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
#Import TfIdfVectorizer from scikit-learn
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.metrics.pairwise import linear_kernel
from data.data_processing import short_desc
import pandas
def recommend(inputs):
# computing similaries
tfidf_vectorizer = TfidfVectorizer(stop_words='english')
data = pandas.read_csv('./data/processed_data.csv')
feature_column = "Complete Project Information"
data[feature_column] = data[feature_column].fillna('')
matrix = tfidf_vectorizer.fit_transform(data[feature_column])
cosine_similarity = linear_kernel(matrix, matrix)
# scoring
final_scores = []
for input in inputs:
# score individual inputs and add the top to a list of candidates
scores = list(enumerate(cosine_similarity[input]))
scores = sorted(scores, key = lambda score: score[1], reverse = True)
final_scores = final_scores + scores[1:11]
# sort the final list of top recommended repos and take the top 10
final_scores = sorted(final_scores, key = lambda score: score[1], reverse = True)
final_scores = final_scores[1:11]
indices = [entry[0] for entry in final_scores]
print()
for input in inputs:
print("Input: " + str(data["Repository Name"][input]) + "(" + short_desc(data["Description"][input], 75))
print("--------------------")
count = 0
for i in indices:
print("Recommendation # " + str(count) + ": " + str(data["Repository Name"][i]) + ": " + short_desc(data['Description'][i], 75))
count = count + 1
return indices