Text Summarization Berbasis Klasterisasi K-Means dan TextRank pada Berita Berbahasa Indonesia

Regita Putri Permata, Aqilah Jihan Nabila, Nadline Virgita Sarungu', Jehan Rizky Abdillah

Abstract


The growth in the volume of digital news in Indonesia drives the need for automated methods to efficiently group and summarize information. This study developed a text mining pipeline that integrated clustering and extractive summarization to summarize Indonesian-language news by topic. From an initial pool of 80,479 articles, 67,370 remained after cleaning; 5,000 were randomly sampled for computational efficiency, and 3,331 documents remained after filtering and were used for the final analysis. Documents were represented using unigram-bigram TF-IDF, dimensionality-reduced via truncated SVD, and clustered using K-Means; the optimal number of clusters was determined by combining the Elbow Method, Silhouette Score, and Davies-Bouldin Index. Each cluster was then summarized using TextRank and evaluated with ROUGE. The analysis identified 11 clusters as optimal, with a Silhouette Score of 0.0675, producing thematically interpretable groupings such as disasters, politics, law, and crime, although the Silhouette Score itself indicates weak cluster separation. The average ROUGE-1, ROUGE-2, and ROUGE-L scores obtained were 0.1938, 0.0743, and 0.1299, respectively, with the best performance observed in clusters with specific, cohesive topics and the lowest in large, heterogeneous clusters. These results indicate an observed association between topic homogeneity and extractive-summary quality; the findings suggest the potential of the proposed pipeline for topic-based extractive summarization of Indonesian news, though this has not yet been benchmarked against a non-clustered TextRank baseline.

Pertumbuhan volume berita digital di Indonesia mendorong kebutuhan metode otomatis untuk mengelompokkan dan meringkas informasi secara efisien. Penelitian ini mengembangkan pipeline text mining yang mengintegrasikan klasterisasi dan extractive summarization untuk merangkum berita berbahasa Indonesia berdasarkan topik. Dari 80.479 artikel awal, tersisa 67.370 artikel setelah pembersihan; 5.000 artikel diambil secara acak untuk efisiensi komputasi, dan 3.331 dokumen tersisa setelah penyaringan serta digunakan pada analisis akhir. Dokumen direpresentasikan dengan TF-IDF kombinasi unigram-bigram dan direduksi dimensinya menggunakan Truncated SVD, lalu dikelompokkan dengan algoritma K-Means; jumlah klaster optimal ditentukan melalui kombinasi Elbow Method, Silhouette Score, dan Davies-Bouldin Index. Setiap klaster kemudian diringkas menggunakan metode TextRank dan dievaluasi dengan metrik ROUGE. Analisis menghasilkan 11 klaster optimal dengan Silhouette Score sebesar 0,0675, menghasilkan pengelompokan tematik yang cukup dapat diinterpretasikan seperti topik bencana, politik, hukum, dan kriminalitas, meskipun nilai Silhouette tersebut menunjukkan pemisahan klaster yang masih lemah. Rata-rata skor ROUGE-1, ROUGE-2, dan ROUGE-L yang diperoleh adalah 0,1938, 0,0743, dan 0,1299, dengan performa terbaik pada klaster bertopik spesifik dan kohesif serta terendah pada klaster besar yang heterogen. Hasil ini menunjukkan adanya asosiasi yang teramati antara homogenitas topik dan kualitas ringkasan ekstraktif; temuan ini mengindikasikan potensi pipeline yang diusulkan untuk extractive summarization berbasis topik pada berita Indonesia, meskipun belum dibandingkan dengan baseline TextRank tanpa klasterisasi.

Keywords


Text summarization; Extractive; K-Means; TextRank; Topik

Full Text:

PDF

References


Chicco, D., Campagner, A., Spagnolo, A., Ciucci, D., & Jurman, G. (2025). The Silhouette coefficient and the Davies-Bouldin index are more informative than Dunn index, Calinski-Harabasz index, Shannon entropy, and Gap statistic for unsupervised clustering internal evaluation of two convex clusters. PeerJ Computer Science, 11, e3309. https://doi.org/10.7717/peerj-cs.3309

Girsang, A. S., & Amadeus, F. J. (2023). Extractive text summarization for Indonesian news article using ant system algorithm. Journal of Advances in Information Technology, 14(2), 295–301. https://www.jait.us/show-227-1320-1.html

Gulati, V., Kumar, D., Popescu, D. E., & Hemanth, J. D. (2023). Extractive article summarization using integrated TextRank and BM25+ algorithm. Electronics, 12(2), 372. https://doi.org/10.3390/electronics12020372

Hidayat, M. Y., Yaqin, M. A., & Abidin, Z. (2025). Semantic-enhanced news clustering using TF-IDF and WordNet with K-Means. Journal of Information Systems and Informatics. https://doi.org/10.63158/journalisi.v7i4.1360

Jain, R., Singh, P., & Puri, S. (2024). Summarization of daily news using TextRank and TF-IDF algorithm. In S. Kumar, B. K., J. H. Kim, & J. C. Bansal (Eds.), Fourth Congress on Intelligent Systems (Vol. 865, pp. 313–324). Springer Nature Singapore. https://doi.org/10.1007/978-981-99-9043-6_26

Jin, K., Paik, S., Yun, J., Jang, S., & Kim, Y. (2024). PRISM: Personalizing reporting with intelligent summarization through multiple frames. IEEE Access. https://doi.org/10.1109/ACCESS.2024.3505612

Juna, M. F., & Hayaty, M. (2023). The observed preprocessing strategies for doing automatic text summarizing. Computer Science and Information Technologies, 4(2), 119–126. https://doi.org/10.11591/csit.v4i2.p119-126

Lin, N., Li, J., & Jiang, S. (2022). A simple but effective method for Indonesian automatic text summarisation. Connection Science, 34(1), 29–43. https://doi.org/10.1080/09540091.2021.1937942

Liu, W., Sun, Y., Yu, B., Wang, H., Peng, Q., Hou, M., Guo, H., Wang, H., & Liu, C. (2024). Automatic text summarization method based on improved TextRank algorithm and K-Means clustering. Knowledge-Based Systems, 287, 111447. https://doi.org/10.1016/j.knosys.2024.111447

Manning, C. D., Raghavan, P., & Schütze, H. (2008). Introduction to information retrieval. Cambridge University Press. https://nlp.stanford.edu/IR-book/

Permata, R., Rendika, & Julianty, L. C. (2025). Towards an Automated Essay Evaluation System NLP Based Text Embeddings and Similarity Metrics. Digital Zone: Jurnal Teknologi Informasi Dan Komunikasi, 16(1), 37–46. https://doi.org/10.31849/digitalzone.v16i1.26541

Purniawan, I. M. A., Sasmita, G. M. A., & Pratama, I. P. A. E. (2022). Clustering berita menggunakan algoritma TF-IDF dan K-Means dengan memanfaatkan sumber data crawling pada situs Detik.com. JITTER: Jurnal Ilmiah Teknologi dan Komputer, 3(1), 821. https://doi.org/10.24843/JTRTI.2022.v03.i01.p18

Putri, I. G. A. I. U., Trisna, I. N. P., & Rusjayanthi, N. K. D. (2025). Abstractive Text Summarization to Generate Indonesian News Highlight Using Transformers Model. Journal of Information Systems and Informatics, 7(2), 1248–1263. https://doi.org/10.51519/journalisi.v7i2.1082

Rifaldi, D., Fadlil, A., & Herman. (2023). Teknik preprocessing pada text mining menggunakan data tweet "Mental Health." Decode: Jurnal Pendidikan Teknologi Informasi, 3(2), 161–171. https://doi.org/10.51454/decode.v3i2.131

Salton, G., & Buckley, C. (1988). Term-weighting approaches in automatic text retrieval. Information Processing & Management, 24(5), 513–523. https://doi.org/10.1016/0306-4573(88)90021-0

Surianto, D. F., & Surianto, D. F. (2024). Enhancing K-Means clustering for journal articles using TF-IDF and LDA feature extraction. Brilliance: Research of Artificial Intelligence, 4(2), 964–972. https://doi.org/10.47709/brilliance.v4i2.5547

Tuhpatussania, S., Utami, E., Hartanto, A. D., et al. (2022). Comparison of LexRank algorithm and maximum marginal relevance in summary of Indonesian news texts online. Pilar Nusa Mandiri, 18(2), 187–192. https://doi.org/10.33480/pilar.v18i2.3190

Wicaksono, A. P., Widjaja, S., Nugroho, M. F., & Putri, C. P. (2024). Elbow and silhouette methods for K value analysis of ticket sales grouping on K-Means. SISTEMASI, 13(1), 28. https://doi.org/10.32520/stmsi.v13i1.2889

Widyassari, A. P., Rustad, S., Shidik, G. F., Noersasongko, E., Syukur, A., Affandy, A., & Setiadi, D. R. I. M. (2022). Review of automatic text summarization techniques & methods. Journal of King Saud University - Computer and Information Sciences, 34(4), 1029–1046. https://doi.org/10.1016/j.jksuci.2020.05.006

Yulianti, E., Pangestu, N., & Jiwanggi, M. A. (2023). Enhanced TextRank using weighted word embedding for text summarization. International Journal of Electrical and Computer Engineering (IJECE), 13(5), 5472. https://doi.org/10.11591/ijece.v13i5.pp5472-5482




DOI: https://doi.org/10.37905/jji.v8i2.40448

Refbacks

  • There are currently no refbacks.