Penerapan Metode Klasifikasi K-Nearest Neighbor dan Lexicon Based untuk Analisis Sentimen Publik Terhadap Pemindahan Ibu Kota Negara dengan Ekstraksi Fitur TF-IDF

Analyze opinions expressed on X (formerly Twitter) regarding the relocation of Indonesia's capital city using combination of classifiers algorithm K-Nearest Neighbor (KNN), Feature Extraction TF-IDF, SMOTE for balancer, and also Lexicon-based approach for labeling data as positive, negative, or neutral sentiment.

Input

• dataset
  • Data regarding the relocation of Indonesia's capital city from X (formerly Twitter).
  • Data crawling was done at 26 October 2024 with keywords: ibu kota baru, ibu kota nusantara, ibu kota pindah, ikn, pemindahan ibu kota, ibukota baru, ibukota nusantara, ibukota pindah, pemindahan ibukota.
  • Some X Search Queries like: lang:id, since:2024-01-01, until:2024-10-01.
  • Also the crawler use LATEST or TOP tab from X.
• slang, stopwords, negation, etc
  • /dictionaries
• Custom Lexicon
  • /lexicons
• Ekstraksi Fitur
  • TF-IDF
• Balancer
  • SMOTE
• Classifier Algorithm
  • K-Nearest Neighbor (KNN)
• Evaluation
  • Confusion Matrix
  • Classification Report

Dependencies

Look at requirements.txt for more details.

Flowchart

/flowcharts

Getting Started

Prerequisites On Machine

• git
• python
• pip
• virtualenv

Cookies

• Export x cookies from browser with some cookies extractor extension on browser.
• Save it to scraper/raw_cookies.json.
• Generate twikit cookies with scraper/cookies.py.
• Save it to scraper/twikit_cookies.json.
• Run scraper/scraper.py.

virtualenv

• Initialize virtualenv

virtualenv <virtualenv_name>

• Activate virtualenv

  • Windows

    <virtualenv_name>/Scripts/activate

  • Linux/macOS

    source <virtualenv_name>/Scripts/activate

• Install dependencies on virtualenv

pip install -r requirements.txt

• Check installed dependencies on virtualenv

pip freeze

• Install a new kernel for Jupyter Notebook named <virtualenv_name> on virtualenv

python -m ipykernel install --user --name <virtualenv_name>

• Update pip

pip install wheel setuptools pip --upgrade

• Start Jupyter Notebook server on virtualenv

jupyter notebook

• Deactivate virtualenv

deactivate

Streamlit

streamlit run streamlit.py