SISTEM PENCARIAN JUDUL SKRIPSI MENGGUNAKAN METODE VECTOR SPACE MODEL BERBASIS WEB
No Thumbnail Available
Date
2022-07
Authors
Journal Title
Journal ISSN
Volume Title
Publisher
Elfitra
Abstract
Every year, student thesis research is increasing and with the addition of these
documents, users will search for thesis titles with appropriate keywords. To make it
easier for users to search for thesis titles, the author makes a system, namely a Vector
Space Model-based thesis title search system which aims to make it easier for users to
find thesis titles recommended by the system by knowing how accurate the search is.
The vector space model method is used to measure the similarity between a document
and a query. The basic concept of the method is to calculate the distance between
documents and then sort them based on the level of proximity.The work principle starts
with case folding, data cleaning, indexing, filtering, stemming, and tokenization, namely
the cutting stage of the input string based on each word that compiles it and splitting
the document into a word frequency table. All words in the document are formed into
one, which is called a term. Each document is displayed as a vector which will be
compared with the terms that have been formed. Lastly, Similarity Analysis to measure
the similarity of documents is done by calculating the Cosine distance between the
documents. The final results of the search for the title of the thesis using the vector
space model method obtained the Cosine Similarity value where D3 has the highest
value with a value of 0.09096, which means that D3 has the highest level of similarity
with the query value. So that the documents that are expected to appear with a high
degree of accuracy can be found precisely and the resulting documents are few.
Description
Keywords
Cosine Similarity, Document, Thesis Title, Query, Vector Space Model
Citation
Perpustakaan