Upload 2 files

app.py

@@ -0,0 +1,106 @@
+from tqdm import tqdm
+from itertools import islice
+from youtube_comment_downloader import *
+from transformers import AutoTokenizer, AutoModelForSequenceClassification, pipeline
+
+import matplotlib.pyplot as plt
+import csv
+import streamlit as st
+import pandas as pd
+import base64
+
+ 
+# Inisialisasi model dan tokenizer
+pretrained= "mdhugol/indonesia-bert-sentiment-classification"
+model = AutoModelForSequenceClassification.from_pretrained(pretrained)
+tokenizer = AutoTokenizer.from_pretrained(pretrained)
+sentiment_analysis = pipeline("sentiment-analysis", model=model, tokenizer=tokenizer)
+label_index = {'LABEL_0': 'positive', 'LABEL_1': 'neutral', 'LABEL_2': 'negative'}
+
+st.title("Youtube Comment Sentimen Analisis")
+
+# Input URL video
+video_url = st.text_input("Masukkan URL video YouTube:")
+
+# Input jumlah komentar yang ingin diambil
+num_comments = st.number_input("Jumlah komentar yang ingin diambil:", min_value=1, value=10)
+
+# Fungsi untuk analisis sentimen
+def analisis_sentimen(text):
+    result = sentiment_analysis(text)
+    label = label_index[result[0]['label']]
+    score = result[0]['score'] * 100
+    return label, score
+
+if st.button("Mulai Analisis"):
+    # Inisialisasi YoutubeCommentDownloader
+    downloader = YoutubeCommentDownloader()
+
+    # Mendapatkan komentar
+    comments = downloader.get_comments_from_url(video_url, sort_by=SORT_BY_POPULAR)
+
+    # Membuka file CSV untuk menulis
+    with open('comments.csv', mode='w', encoding='utf-8', newline='') as file:
+        # Membuat objek writer
+        writer = csv.DictWriter(file, fieldnames=['cid', 'text', 'time', 'author', 'channel', 'votes', 'photo', 'heart', 'reply'])
+        
+        # Menulis header
+        writer.writeheader()
+        
+        # Menulis data komentar
+        for comment in tqdm(islice(comments, num_comments)):
+            # Menghapus kolom 'time_parsed' dari komentar
+            comment.pop('time_parsed', None)
+            writer.writerow(comment)
+
+    st.success(f"Komentar berhasil diunduh dan disimpan dalam file 'comments.csv'")
+    
+    # Membaca data dari file CSV
+    comments_df = pd.read_csv('comments.csv')
+    
+    #analisis sentimen
+    st.info("Memulai analisis sentimen....")
+
+    # List untuk menyimpan hasil analisis sentimen
+    hasil_analisis = []
+
+    # Membaca data dari file CSV
+    with open('comments.csv', mode='r', encoding='utf-8') as file:
+        reader = csv.DictReader(file)
+        for row in tqdm(reader):
+            comment_text = row['text']
+            label, score = analisis_sentimen(comment_text)
+            hasil_analisis.append((comment_text, label, score))
+
+    # Menampilkan hasil analisis sentimen
+    st.subheader("Hasil Analisis Sentimen")
+    #st.write(hasil_analisis)
+
+    # Menampilkan histogram
+    labels, scores = zip(*[(label, score) for _, label, score in hasil_analisis])
+    plt.hist(labels, bins=30, color='blue', alpha=0.7, edgecolor='black')
+    plt.xlabel('Skor Sentimen')
+    plt.ylabel('Jumlah Komentar')
+    plt.title('Distribusi Sentimen Komentar')
+    st.pyplot(plt)
+
+    # Menghitung jumlah dan persentase
+    jumlah_positif = labels.count('positive')
+    jumlah_negatif = labels.count('negative')
+    jumlah_netral = labels.count('neutral')
+    total_komentar = len(labels)
+    persentase_positif = (jumlah_positif / total_komentar) * 100
+    persentase_negatif = (jumlah_negatif / total_komentar) * 100
+    persentase_netral = (jumlah_netral / total_komentar) * 100
+
+    st.write(f"Total Komentar: {total_komentar}")
+    st.write(f"Persentase Komentar Positif: {persentase_positif:.2f}% / {jumlah_positif} Komentar")
+    st.write(f"Persentase Komentar Negatif: {persentase_negatif:.2f}% / {jumlah_negatif} Komentar")
+    st.write(f"Persentase Komentar Netral: {persentase_netral:.2f}% / {jumlah_netral} Komentar")
+    
+    # Menampilkan tabel dengan menggunakan st.table()
+    st.subheader("Data Komentar")
+    st.table(comments_df)
+    
+   
+

requirements.txt

@@ -0,0 +1,5 @@
+youtube-comment-downloader
+tqdm
+torch
+transformers
+matplotlib