Quick Start with New Data

From TechWiki

In first setting up or expanding an OSF (open semantic framework) instance, configuring for and loading data is always a major component. This Quick Start guide is meant to be a short supplement to the more detailed dataset workflows.

Contents 1 Step 1: Map Out Game Plan 2 Step 2: Enhance or Extend Ontology 3 Step 3: Prepare and Load Datasets 4 Step 4: Refine and Care and Feed Ontology

Step 1: Map Out Game Plan

• Read essential background materials about datasets; key background documents are listed below
• Define datasets
  • See the Datasets: Identify document
• Define dataset attributes and structure
  • See the Datasets: Analyze and Structure Requirements document
  • See the Datasets: Define Dataset Specifications document
• Review existing ontology
  • Become familiar with (initially) the Protégé ontology management tool
  • Become familiar with the existing ontology (should one exist) with structOntology^[1] and the relation browser
• Plan out approach
  • Define needed concepts (classes) and attributes (properties) in relation to the datasets anticipated.

Step 2: Enhance or Extend Ontology

Note: You may be best to use Protégé at this stage, since many of the ontology changes may be large or in bulk. See further structOntology v Protégé?

• Add your new ontology classes
  • See the Adding an Ontology Concept using Protégé document
• Add your new properties or attributes
  • See the section on Adding a New Attribute
• Test your extended ontology
  • See for the using reasoners with Protégé document.

Step 3: Prepare and Load Datasets

Note: this step assumes you will be loading commON datasets; see the large datasets discussion for that alternate path

• Be familiar with the irON specification, specifically the commON (CSV) specification
  • You may also want to check out the CommON Case Study
• Prepare up your full datasets, including linkage schema
  • See the Datasets: Prepare Small (commON) Datasets document
• Submit your datasets to the conStruct Import module
  • See the Datasets: Import Dataset Files document
• If errors are shown at time of import, revise and re-import.

Step 4: Refine and Care and Feed Ontology

Note: You may be best to use structOntology for ongoing refinements. See further structOntology v Protégé?

• You are encouraged to add to the metadata and altLabels (especially) for your existing ontology as you work with it. structOntology is the proper tool for this job.

1. ↑ See further the Individual conStruct Ontology (structOntology) Tool manual.