skip to main content

Sistem Temu Kembali Informasi dengan Metode Vector Space Model

*Fatkhul Amin  -  Fakultas Teknologi Informasi, Universitas Stikubank, Semarang, Indonesia

Citation Format:
Abstract

The information retrieval system (IRS) now provides  document  search results  with abundant results of  documents  (high recall) and low accuracy  (low  precision).   The  objective  of  designing  IRS  with  Vector  Space  Model  (VSM)  Method  is  to  facilitate  users  to  searchIndonesian  documents.  IRS  Software  is  designed  to  provide  search  results  with  the  optimum  number  of  documents  (low  recall)  and accuracy (high precision) with VSM method  that users  may  get fast and accurate  results.  VSM  method  provides a different  credit  for each  document  stored in  a  database  which in turns to determine  the  document most similar  to  the query,  where the documents with the highest credits are placed  on the top of  the  search results.  The evaluation  of  search results  with  IRS  is conducted under  recall  andprecision tests.  This study  fascinatingly creates  a system which can preprocess  (tokenizing,  filtering,  and  stemming)  within  computation time  of four  minutes  forty-one seconds.  The system  can  search  the  documents  and  display  them in  the search results  with an averagecomputation time of 1.5 seconds, with  an average recall of 0.19, and an average precision of 0.54.  The system is equipped with a credit for each document, and the value is positioned on which a user may easily search the Indonesian text.

Keywords: IRS, Vector Space Model, Recall, Precision

Note: This article has supplementary file(s).

Fulltext |  Research Instrument
Untitled
Subject
Type Research Instrument
  Download (684KB)    Indexing metadata

Article Metrics:

Last update:

  1. The Stemming Application on Affixed Javanese Words by using Nazief and Adriani Algorithm

    Dewi Soyusiawaty, Anna Hendri Soleliza Jones, Nora Lestari Lestariw. IOP Conference Series: Materials Science and Engineering, 771 (1), 2020. doi: 10.1088/1757-899X/771/1/012026

Last update: 2024-03-28 13:48:55

No citation recorded.