OBOE An ontology for describing synthesizing ecological data

OBOE: An ontology for describing & synthesizing ecological data Knowledge Representation Working Group

Ecological research • Research in ecology increasing relies on the synthesis of data (physical, chemical and biological) • Problem: data are heterogeneous; details not recorded • Metadata standards important first step, but don’t capture all necessary aspects of data content • Solution: map data and metadata to a formal model that captures their meaning (or semantics)

Motivation Formal information is needed to: • Comprehensively discover data – Search for and access relevant data – Rapidly interpret, summarize and view data • Automatically integrate data – Automatically determine if data are compatible – Calculate appropriate conversions to merge data

Definitions • Observation: – An assertion of the existence of an entity, by an observer (human or non-human), typically resulting in one or more measurements of characteristics of that entity. – Observations can provide context for other observations. – Entities can be biotic (e. g. , animals) and abiotic (e. g. , water) • Observational data: – Any recorded measurements resulting from observations

Observation ? A assertion that an entity exists

Entity All things concrete and conceptual

Entity An extension point for domain-specific terms

Measurement Observations can result in measurements of characteristics of the entity

Characteristic

Measurement The raw data Measurement assigns a value, via a measurement standard, to the characteristic

Measurement standard All the units, scales, indices, classifications, and lists used for ‘measuring’ a characteristic

Example Measuring the height (characteristic) in meters (standard) of an tree (entity)

Context Observations can provide context for other observations

Context is transitive; measurements can be made at each level of observation

Model extensibility • OBOE provides a core framework for organizing domain concepts – • Entities, Characteristics and Measurement Standards Developing extensions – – – Units Top-level ecological concepts (textbook parsing) Structured controlled vocabularies (LTER)

Semantic annotation

Semantic annotation Example data set: the abundance of Trapeziid crabs in coral colonies (Stewart et al. 2006)

Semantic annotation

Semantic annotation

Semantic annotation

Semantic annotation

Applications overview • • • Core OBOE ontology definitions (complete) Semantic annotation mechanism (prototype) Visualization of observational structures (prototype) Semantic search and ranking (prototype) Automated data summarization (development) Data integration (active research & development)

Semantic annotation

Simple Keyword Search

Broader architecture EML Semantic Annotation Scientists & other end users Metadata Editing Data Discovery Data Browsing Creation & Management of Standard Ontologies Community-Driven (Collaboration Among Scientists & Informatics Specialists) Federated Metadata & Data Management Back-End System Scientists & other end users Data Discovery Data Browsing

Summary • OBOE is an ontology framework for describing observations of entities, their measurements, and context • OBOE provides a structured approach for incorporating domain ontologies • OBOE is used to semantically annotate observational data • OBOE provides necessary constructs for discovering and integrating the diverse range of data

Acknowledgements • Knowledge Representation Working Group • Mark Schildhauer, Matt Jones (NCEAS) • Shawn Bowers, Bertram Ludaescher, Dave Thau (UCD) • Deana Pennington (UNM) • Serguei Krivov, Ferdinando Villa (UVM) • Rich Williams (Microsoft)