metadata

license: apache-2.0
datasets:
  - ZakyF/PRDECT-ID
language:
  - id
metrics:
  - accuracy
evaluation:
  - task:
      type: text-classification
      name: Sentiment Analysis
    metrics:
      - name: Accuracy
        type: accuracy
        value: 1
      - name: Cross-Validation Accuracy
        type: accuracy
        value: 0.99981
pipeline_tag: text-classification
library_name: sklearn
tags:
  - sentiment-analysis
  - nlp
  - naive-bayes
  - e-commerce
  - indonesian

Sentiment Analysis

Model SVM dan Naive Bayes untuk mengklasifikasikan ulasan ke dalam kategori Bagus, Normal, atau Buruk menggunakan PRDECT-ID Dataset.

Deskripsi

Model ini menganalisis ulasan pelanggan Tokopedia untuk menghasilkan insight seperti rekomendasi perbaikan pengiriman atau kualitas produk.

Penggunaan

import pickle
from sklearn.preprocessing import LabelEncoder, StandardScaler

# Load model dan preprocessing
svm_model = pickle.load(open('svm_model.pkl', 'rb'))
scaler = pickle.load(open('scaler.pkl', 'rb'))
le_sentiment = pickle.load(open('le_sentiment.pkl', 'rb'))
le_emotion = pickle.load(open('le_emotion.pkl', 'rb'))

# Contoh prediksi
data = [[5, 'Positive', 'Happy']]  # Rating, Sentiment, Emotion
data_scaled = scaler.transform(data)
prediksi = svm_model.predict(data_scaled)
print(prediksi)  # Output: ['Bagus']