Decision Support System using Weighting Similarity Model for Constructing Ground-Truth Dataset

This research aims to form a ground-truth dataset in the entity-matching process used to detect duplication of records in a bibliographic database. The contribution of this research is the obtained dataset which can be used as reference in measuring and evaluating the entity matching model implemented in bibliographic databases. This aim was achieved by developing a decision support system through experts who act as decision makers in the bibliographic databases field to construct ground-truth datasets. The model used in this decision support system weights similarity by comparing each attribute of the pairwise record in the dataset. An expert who understands all characteristics of the research database can use the graphical user interface to evaluate and determine the pairwise record that meets the conditions, such as duplication of records. This research produces a ground-truth dataset using the decision support system approach.