Please use this identifier to cite or link to this item:
|Title:||Comparative weighting methods of vector space model|
Sakda Arj Ong Vallipakorn
King Mongkut's University of Technology North Bangkok
|Citation:||ARPN Journal of Engineering and Applied Sciences. Vol.10, No.3 (2015), 1316-1323|
|Abstract:||© 2006-2015 Asian Research Publishing Network (ARPN). This research aimed to develop a program for data retrieval stored in the form of questions and answers. Fidility Fascinate Fastness Co., Ltd., Thailand has been used an old traditional of storage and retrieval of knowledge system in the form of Google Drive, which was inconvenient and time consuming when retrieving the desired knowledge. Therefore, the new development of knowledge retrieval based on Vector Space Model (VSM) to facilitate the users in the knowledge retrieval was conducted and invented to solve the problems. For VSM concept, the required knowledge from the database was transformed by with C# and wrapped by the Longtext Matching, then indexed cutting by Inverted Indexing Search. Information retrieval and sorting results was robustness based on algorithm of VSM. The results of knowledge retrieval of 200 questions were processed by 100 queries. The Cosine formula shows the best appropriated formula than Dice and Jaccard formulas which return the higher of their precision (82.50 %), recall values (97.35%), and accuracy (89.31%) measured by F-measurement.|
|Appears in Collections:||Scopus 2011-2015|
Files in This Item:
There are no files associated with this item.
Items in DSpace are protected by copyright, with all rights reserved, unless otherwise indicated.