prosekutor commited on
Commit
d799ebd
·
1 Parent(s): 2d7e826

added application files

Browse files
Files changed (3) hide show
  1. app.py +68 -0
  2. models/model_v1.pkl +3 -0
  3. models/transformer_v1.pkl +3 -0
app.py ADDED
@@ -0,0 +1,68 @@
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
+ import gradio as gr
2
+ import joblib as jb
3
+ import re
4
+ import string
5
+ from nltk.corpus import stopwords
6
+ from nltk.tokenize import word_tokenize
7
+ from nltk.stem import PorterStemmer, WordNetLemmatizer
8
+ import nltk
9
+ nltk.download('stopwords')
10
+ stop_words = set(stopwords.words('english'))
11
+
12
+ cv = jb.load('.//models//transformer_v1.pkl')
13
+ model = jb.load('.//models//model_v1.pkl')
14
+
15
+
16
+ def preprocess_text(text):
17
+ """
18
+ Runs a set of transformational steps to
19
+ preprocess the text of the tweet.
20
+ """
21
+ # convert all text to lower case
22
+ text = text.lower()
23
+
24
+ # remove any urls
25
+ text = re.sub(r'http\S+|www\S+|https\S+', "", text, flags=re.MULTILINE)
26
+
27
+ # replace '****' with 'curse'
28
+ text = re.sub(r'\*\*\*\*', "gaali", text)
29
+
30
+ # remove punctuations
31
+ text = text.translate(str.maketrans("", "", string.punctuation))
32
+
33
+ # remove user @ references and hashtags
34
+ text = re.sub(r'\@\w+|\#', "", text)
35
+
36
+ # remove useless characters
37
+ text = re.sub(r'[^ -~]', '', text)
38
+
39
+ # remove stopwords
40
+ tweet_tokens = word_tokenize(text)
41
+ filtered_words = [word for word in tweet_tokens if word not in stop_words]
42
+
43
+ # stemming
44
+ ps = PorterStemmer()
45
+ stemmed_words = [ps.stem(w) for w in filtered_words]
46
+
47
+ # lemmatizing
48
+ lemmatizer = WordNetLemmatizer()
49
+ lemma_words = [lemmatizer.lemmatize(w, pos='a') for w in stemmed_words]
50
+
51
+ return ' '.join(lemma_words)
52
+
53
+
54
+ def sentiment_analysis(text):
55
+ print(text)
56
+ text = cv.transform([preprocess_text(text)])
57
+ pred_prob = model.predict_proba(text)[0]
58
+ output = {"Negative": float(pred_prob[0]), "Neutral": float(pred_prob[1]), "Positive": float(pred_prob[2])}
59
+ print(output)
60
+ return output
61
+
62
+
63
+ demo = gr.Interface(
64
+ fn=sentiment_analysis,
65
+ inputs=gr.Textbox(label="Input here", lines=2, placeholder="Input your text"),
66
+ outputs=gr.Label(label="Sentiment Analysis"),
67
+ )
68
+ demo.launch()
models/model_v1.pkl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:28b989f2dec37cb7615dee7d299189decf1d15647fa92e1e9b6f38966a994449
3
+ size 419236163
models/transformer_v1.pkl ADDED
@@ -0,0 +1,3 @@
 
 
 
 
1
+ version https://git-lfs.github.com/spec/v1
2
+ oid sha256:9fb9484ac5da542395636d3f2bed85ee06109f66eb4923d11bc0da05218a8357
3
+ size 232123