Standardization of scientific experimental data representation through ontology-based metadata schema

Halal is a wide area that involved multidisciplinary domains such as biotechnology and medical science in which data and information come from various sources such as laboratory instruments and machines. Halal is defined as the status of certain products that do not contain unpermitted ingredients. The halal determination for various samples and ingredients were done using various laboratory instrument, at which each instrument has a different structure and format of data. These make it difficult for managing and integrating the data for analysis. Research areas involved with data management and integration need to explore data standardization as it helps in bringing data into a common format. Hence, it would help in collaborative research as well as sharing of data and information. The problem addressed in this study is, researchers in the determination of Halal components of products require data standardization as it is hard in managing and analyzing scientific experimental data from multiple laboratory instruments that have different structures and formats of data. The objective of this research was to standardize scientific experimental data from Halal Institute laboratory instruments. To accomplish such goals, an ontology schema model was proposed to standardize and gives a controlled vocabulary of the scientific experimental data from Halal Institute laboratory instruments. A metadata representation structure, based on the proposed ontology schema, was also developed to give a standard structure to the scientific experimental data representation as well as simplified data that enables data retrieval and display. Two types of evaluation were conducted in this study which was; ontology schema evaluation and metadata representation structure evaluation. Both evaluations were done using the data files from instruments in the laboratory for raw and processed materials and liquid analysis, namely Gas Chromatography-Mass Spectrometry (GCMS) and High Performance Liquid Chromatography (HPLC) instruments. The proposed ontology schema model was evaluated and validated based on completeness and correctness analysis measures. It was to ensure that the proposed ontology schema model was designed completely and correctly based on the grouped data and information from the laboratory instruments. Based on the ontology schema model evaluation, the completeness percentage of the ontology schema model was 100%, conform to all of the grouped data Sample Info, Result Info, Experimental Setup Info, and Graph Info from the laboratory instruments. For the correctness percentage, the result of the ontology schema model correctly conforming to the Sample Info data of GC-MS and HPLC instruments which were 50% and 43% respectively. The correctness percentage conforms to the Result Info and Graph Info data of both instruments were 100%. For the correctness percentage conformed to the Experimental Setup Info data of GC-MS and HPLC instruments, was 96% and 86% respectively. These figures indicate that the average recall percentage of the IEDOS correctly conforms to all of the grouped data was 84%. Overall, the results gained were satisfactory although the results of the correctness percentage conform to Sample Info data was slightly lower because of data selection factors. Metadata representation structure evaluation and validation consists of precision and recall analysis to measure the accuracy of metadata extraction from the laboratory instruments data files. The precision percentages were 90% and recall were 100% for both GCMS and HPLC instruments data files. The results gained shows the appropriate applicability of the proposed ontology-based metadata, in giving a standardized structure for the scientific experimental data for these instruments. This could positively facilitate the analysis of the scientific experimental data by giving the same structure of data to be compared and evaluated.

Text IPPH 2020 1 IR.pdf Download (1MB)

Actions (login required)

View Item