Commit
Β·
d6934dd
1
Parent(s):
28ddb3f
update space
Browse files- .gitignore +1 -0
- app.py +83 -0
- notebook/sentimen_analysis_brimo.ipynb +0 -0
- random_forest_model.pkl +3 -0
- requirements.txt +7 -0
- tfidf_vectorizer.pkl +3 -0
.gitignore
ADDED
@@ -0,0 +1 @@
|
|
|
|
|
1 |
+
venv
|
app.py
ADDED
@@ -0,0 +1,83 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Import library yang diperlukan
|
2 |
+
import streamlit as st
|
3 |
+
import joblib
|
4 |
+
import re
|
5 |
+
import string
|
6 |
+
import nltk
|
7 |
+
from nltk.corpus import stopwords
|
8 |
+
from nltk.tokenize import word_tokenize
|
9 |
+
from Sastrawi.Stemmer.StemmerFactory import StemmerFactory
|
10 |
+
|
11 |
+
# Download data NLTK
|
12 |
+
nltk.download("stopwords")
|
13 |
+
nltk.download("punkt")
|
14 |
+
nltk.download("punkt_tab")
|
15 |
+
|
16 |
+
# Load model dan vectorizer
|
17 |
+
model = joblib.load("random_forest_model.pkl")
|
18 |
+
vectorizer = joblib.load("tfidf_vectorizer.pkl")
|
19 |
+
|
20 |
+
# Stemmer Bahasa Indonesia
|
21 |
+
factory = StemmerFactory()
|
22 |
+
stemmer = factory.create_stemmer()
|
23 |
+
|
24 |
+
|
25 |
+
# Fungsi untuk membersihkan teks dari karakter yang tidak diperlukan
|
26 |
+
def delete_unused_char(text):
|
27 |
+
text = re.sub(r"@[A-Za-z0-9]+", "", text) # menghapus mention
|
28 |
+
text = re.sub(r"#[A-Za-z0-9]+", "", text) # menghapus hashtag
|
29 |
+
text = re.sub(r"RT[\s]", "", text) # menghapus RT
|
30 |
+
text = re.sub(r"http\S+", "", text) # menghapus link
|
31 |
+
text = re.sub(r"[0-9]+", "", text) # menghapus angka
|
32 |
+
text = re.sub(r"[^\w\s]", "", text) # menghapus karakter selain huruf dan angka
|
33 |
+
text = text.replace("\n", " ") # mengganti baris baru dengan spasi
|
34 |
+
text = text.translate(
|
35 |
+
str.maketrans("", "", string.punctuation)
|
36 |
+
) # menghapus semua tanda baca
|
37 |
+
text = text.strip(" ") # menghapus karakter spasi dari kiri dan kanan teks
|
38 |
+
return text
|
39 |
+
|
40 |
+
|
41 |
+
# Fungsi untuk membersihkan teks
|
42 |
+
def cleaned_text(text):
|
43 |
+
delete_unused_char(text)
|
44 |
+
# 1. Lowercasing
|
45 |
+
text = text.lower()
|
46 |
+
# 2. Remove punctuation
|
47 |
+
text = text.translate(str.maketrans("", "", string.punctuation))
|
48 |
+
# 3. Remove numbers
|
49 |
+
text = re.sub(r"\d+", "", text)
|
50 |
+
# 4. Tokenization
|
51 |
+
words = word_tokenize(text)
|
52 |
+
# 5. Remove stopwords
|
53 |
+
stop_words = set(stopwords.words("indonesian")) # Stopwords bahasa Indonesia
|
54 |
+
words = [word for word in words if word not in stop_words]
|
55 |
+
# 6. Ubah kate ke bentu asli dengan Stemmer Sastrawi
|
56 |
+
words = [stemmer.stem(word) for word in words]
|
57 |
+
return " ".join(words)
|
58 |
+
|
59 |
+
|
60 |
+
# Fungsi untuk prediksi sentimen
|
61 |
+
def predict_sentiment(text):
|
62 |
+
text = cleaned_text(text) # Preprocessing sebelum prediksi
|
63 |
+
X = vectorizer.transform([text]) # Ubah teks menjadi vektor
|
64 |
+
prediction = model.predict(X)[0] # Prediksi sentimen
|
65 |
+
return prediction
|
66 |
+
|
67 |
+
|
68 |
+
# Streamlit UI
|
69 |
+
st.title("Analisis Sentimen Review BRI Mobile π³")
|
70 |
+
st.write("Masukkan review dan dapatkan prediksi sentimen (Positif, Negatif, Netral)")
|
71 |
+
|
72 |
+
# Input review dari pengguna
|
73 |
+
user_input = st.text_area("Masukkan review di sini:")
|
74 |
+
|
75 |
+
if st.button("Prediksi Sentimen"):
|
76 |
+
if user_input.strip() == "":
|
77 |
+
st.warning("Silakan masukkan teks terlebih dahulu!")
|
78 |
+
else:
|
79 |
+
sentiment = predict_sentiment(user_input)
|
80 |
+
st.success(f"Prediksi Sentimen: **{sentiment}**")
|
81 |
+
# st.write(cleaned_text(user_input))
|
82 |
+
|
83 |
+
st.write("Dibuat dengan π oleh Muhammad Farkhan Adhitama")
|
notebook/sentimen_analysis_brimo.ipynb
ADDED
The diff for this file is too large to render.
See raw diff
|
|
random_forest_model.pkl
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:f1a6a774ffd94d3a66550a24d02862caa27e85c711258d07a127baf00f158833
|
3 |
+
size 65195937
|
requirements.txt
ADDED
@@ -0,0 +1,7 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
streamlit
|
2 |
+
scikit-learn
|
3 |
+
joblib
|
4 |
+
nltk
|
5 |
+
numpy
|
6 |
+
pandas
|
7 |
+
Sastrawi
|
tfidf_vectorizer.pkl
ADDED
@@ -0,0 +1,3 @@
|
|
|
|
|
|
|
|
|
1 |
+
version https://git-lfs.github.com/spec/v1
|
2 |
+
oid sha256:65eabfb11ec6e48408db1619a37d01719ac5781385d4720582e451b6f87aa2d4
|
3 |
+
size 8008
|