{
      "thesis_name": "Sarjana",
      "department": "Fakultas Matematika dan Pengetahuan Alam",
      "userid": 46,
      "date_type": "published",
      "metadata_visibility": "show",
      "lastmod": "2026-07-29 01:51:17",
      "corp_creators": [
        "Universitas Pakuan",
        "Fakultas Matematika dan Ilmu Pnegetahuan Alam",
        "Program Studi Ilmu Komputer"
      ],
      "creators": [
        {
          "name": {
            "honourific": null,
            "lineage": null,
            "family": "Fadhilah",
            "given": "Anas"
          },
          "NPM": 65121021
        }
      ],
      "datestamp": "2026-07-29 01:51:17",
      "abstract": "Social media has become a major platform for expressing public opinion regarding social, political, and\r\neconomic issues. One of the viral phenomena currently discussed in Indonesia is the hashtag #KaburAjaDulu,\r\nwhich reflects public concerns regarding employment opportunities, economic instability, and social\r\nconditions. This study aims to compare the performance of Support Vector Machine (SVM) combined with\r\nWord2Vec and Bidirectional Encoder Representations from Transformers (BERT) in sentiment analysis of\r\ntweets containing the #KaburAjaDulu hashtag on X (Twitter). The dataset was collected using web scraping\r\ntechniques and processed through several preprocessing stages, including cleaning text, case folding, slang\r\nnormalization, tokenization, stopword removal, and sentence reconstruction. The research adopted the Cross\r\nIndustry Standard Process for Data Mining (CRISP-DM) framework consisting of business understanding,\r\ndata understanding, data preparation, modeling, evaluation, and deployment stages. In the first approach,\r\nWord2Vec with the Skip-gram architecture was used to generate word embeddings which were classified\r\nusing SVM with a Radial Basis Function (RBF) kernel. The second approach employed IndoBERT for\r\ncontextual sentiment classification. Model evaluation was conducted using accuracy, precision, recall, F1-\r\nscore, and confusion matrix metrics. Experimental results demonstrated that BERT achieved superior\r\nclassification performance compared with SVM combined with Word2Vec due to its contextual\r\nunderstanding capability. The findings indicate that transformer-based models provide more accurate\r\nsentiment classification for Indonesian social media text, especially in detecting contextual and implicit\r\nsentiment expressions.\r\nKeywords: Sentiment Analysis; Support Vector Machine; Word2Vec; BERT; Twitter; Natural Language\r\nProcessing; CRISP-DM; #KaburAjaDulu",
      "eprint_status": "archive",
      "uri": "http:\/\/eprints.unpak.ac.id\/id\/eprint\/11071",
      "institution": "Universitas Pakuan",
      "subjects": [
        "QK"
      ],
      "divisions": [
        "sch_ecs"
      ],
      "eprintid": 11071,
      "type": "thesis",
      "dir": "disk0\/00\/01\/10\/71",
      "thesis_type": "Skripsi",
      "rev_number": 6,
      "date": "2025-06-21",
      "status_changed": "2026-07-29 01:51:17",
      "full_text_status": "none",
      "contributors": [
        {
          "name": {
            "honourific": null,
            "lineage": null,
            "family": "Wihartiko",
            "given": "Fajar Delli"
          },
          "type": "http:\/\/www.loc.gov\/loc.terms\/relators\/THS"
        },
        {
          "name": {
            "honourific": null,
            "lineage": null,
            "family": "Andriani",
            "given": "Siska"
          },
          "type": "http:\/\/www.loc.gov\/loc.terms\/relators\/THS"
        }
      ],
      "title": "Analisis Sentimen Pada Hashtag #Kabupaten Aja dulu Menggunakan Metode Support Vector Machine (SVM) Dengan Word2vec Bidirectional Encoder Representations From Transformers (BERT)",
      "ispublished": "pub"
    }