1st International Conference on Science and Technology, ICOST 2019, 2-3 May, Makassar, Indonesia

Research Article

Computing Group News Documents Using K-Means and K-Nearest Neighbor

Download771 downloads
  • @INPROCEEDINGS{10.4108/eai.2-5-2019.2284616,
        author={Sitti  Arni and Syaharullah  Disa},
        title={Computing Group News Documents Using K-Means and K-Nearest Neighbor},
        proceedings={1st International Conference on Science and Technology, ICOST 2019, 2-3 May, Makassar, Indonesia},
        publisher={EAI},
        proceedings_a={ICOST},
        year={2019},
        month={6},
        keywords={classification clustering information retrieval},
        doi={10.4108/eai.2-5-2019.2284616}
    }
    
  • Sitti Arni
    Syaharullah Disa
    Year: 2019
    Computing Group News Documents Using K-Means and K-Nearest Neighbor
    ICOST
    EAI
    DOI: 10.4108/eai.2-5-2019.2284616
Sitti Arni1,*, Syaharullah Disa2
  • 1: Department of Information Systems, Sekolah Tinggi Manajemen Informatika Komputer Profesional Makassar, Indonesia. 90233
  • 2: Department of Information Engineering, Sekolah Tinggi Manajemen Informatika Komputer AKBA Makassar, Indonesia. 90245
*Contact email: arnist1903@gmail.com

Abstract

The aim of the study was to reduce the computational time of the document search by using the K-Means and K-Nearest Neighbor algorithm. The search result through search engines with display the documents according to the keywords entered. The number of documents that were displayed will make the user difficult to fine the required documents and it a long time for the display process of the documents. K-Means algorithm is used for the clustering documents obtained through online with search engine whereas algorithm K-Nearst Neighbor is used for grouping the clustering result document that are done with offline. The clustering with K-Means can reduce computational time on the news grouping by using K-Nearst Neighbor. The combination of two methods result an average time of 0.5011 seconds. Whereas the grouping that uses pure K-Nearst Neighbor requires the computational time 2.4841 seconds. The test result indicated that the combination of the two algorithms resulted accurate classification with the faster computation time.