PERANCANGAN SISTEM PENDUKUNG KEPUTUSAN PEMILIHAN DOSEN PELAKSANA KERJA SAMA PADA BIRO KEMITRAAN ITB MENGGUNAKAN METODE DATA MINING

• Main Author: Muhammad Alfarisi, Zidane
• Format: Final Project
• Language: Indonesia
• Online Access: https://digilib.itb.ac.id/gdl/view/86088
• Institution: Institut Teknologi Bandung

Bandung Institute of Technology (ITB) always strives to improve partnerships to expand the partnership network at ITB. To realize this goal, there is one of ITB's work units, namely the Partnership and Strategic Management Section under the ITB Partnership Bureau. One of the problems faced in this section is related to achieving the target number of collaborations. On the other hand, ITB has many lecturers who have the potential to be implementers of collaborations. The problem occurs because there is no more structured processing to be able to produce information about the expertise of ITB lecturers. At this time, the determination of implementing collaboration lecturers is carried out inefficiently because it is still done manually and sporadically. This study aims to design a Decision Support System (DSS) for the process of selecting implementing collaboration lecturers efficiently. The method used in designing the SPK follows the prototyping method with the stages of planning, analysis, design, and implementation. Meanwhile, in designing the data mining model, the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework is used. The flow of this research generally begins with mapping existing conditions, identifying needs, modeling, designing subsystem components and DSS prototypes, to evaluating prototypes and making conclusions. This research produces a Decision Support System (DSS) based on publication data sourced from the Scopus database to support the process of selecting implementing lecturers for collaboration at the ITB Partnership Bureau. This system uses data mining models, such as the Latent Dirichlet Allocation (LDA) weighted moving average model, the Yet Another Keyword Extractor (YAKE) model, and network analysis. The designed DSS includes a user interface prototype with the main functions of searching for lecturers and mapping lecturer expertise topics, as well as a data management subsystem component that produces an ERD with six entities.