Automated MIAPPE Compliance Validation

<b>Objective</b>This poster describes the preliminary results of a study of the use of semantic web technologies to tackle both the representation and validation of conformity of Plant Phenotyping Experiment datasets with the MIAPPE standard.<br><b>Motivation</b> Plant phenotyping research generates large datasets that are often not reusable due to the lack of standardization in their descriptors. The Minimal Information about a Plant Phenotyping Experiment (MIAPPE) is a Minimum Information (MI) standard for plant phenotyping, that has been introduced to mitigate this issue. To effect, MIAPPE provides a closed set of metadata descriptors which a plant phenotyping dataset must conform to, in order to be classified as MIAPPE-compliant.<br><br>As researchers begin to use MIAPPE descriptors to annotate their datasets, it will be necessary to assert the compliance of these datasets with the MIAPPE standard in an automated manner, so that they can be inserted into MIAPPE repositories or feedback can be provided to enable researchers to adjust their data annotation to ensure acceptance by such repositories. <br><b>Proposed Approach</b>1. Analyse the current state of both MIAPPE and a published plant phenotyping dataset, and attempt to represent them both under a common data model using semantic web technologies. 2. Attempt to verify the conformation of the metadata descriptors of the plant phenotyping dataset with the MIAPPE specification. 3. Provide documentation of the results of the verification in the form of a report, detailing the extent of compliance of the plant phenotyping dataset with the MIAPPE specification.<b><br></b><b>Current Status</b>Our proposed approach is represented in the BPMN business process attached. At present our effort has focused on the Dataset Representation part of our business process. So far, we’ve been able to successfully pre-process and structure a plant phenotyping dataset, and are in the process of creating the mappings to our selected domain ontology (Plant Phenotyping Experiment Ontology).<br><b>Future Ideas</b>Use similarity matching algorithms to aid in the mapping creation. Explore the possible usage of machine-learning techniques both in the mapping creation as well as the Dataset Validation<br><br><br>