| Summary: | Retrieval models and techniques can be applied to retrieve memo information and
it does relate to certain queries or concepts. Different method of retrieval shows different
relevant documents. Therefore, different methods are used to retrieve, structure and
similarity measures being used to retrieve memo. The data are collected at Center of
HND Studies at TATi University College. The process that has been applied is digitizing,
stop word removal, stemming and building index. The results of this process then stored
in database. In this study 50 document of memo being used, where refer to the titles and
the document. In order to show the measure of different memo structure three similarity
measures being used. The structure refers to the titles of memo, the documents and
combination of both title and documents. The results shows that title structure have the
lowest performance of result compared to document structure and the combination of
both. Meanwhile, in similarity measures the results shows that weighting schemes using
cosine perform well compared to Dice and Russell-Rao. The overall result shows that the
best performance for retrieving memo is by using document structure coupled with
Cosine similarity. (Author's abstract)
|
|---|