Structuring the biodiversity informatics community at the European level and beyond

Service Networks, Service Sets and Biodiversity Catalogue

Alex HardistyCardiff University

Outline

Provision of data and processing of information are both “services”. How can we establish a landscape in which services are both discoverable and interoperable?

• What is a Service Network and what are Service Sets?

• Why do we need them and how do we build them?

• What is the role of the Biodiversity Catalogue?

• Where is this all going?

What is a Service Network?

• A set of Web service (WS) instances that interact together to perform an application objective– In our case: multiple objectives, varying over time and

from one user to another

• Usage and hence composition needs to be dynamic

• In a Service Network:– Instances may join and leave– Instances are discoverable– Managed to a greater or lesser extent

WS1 WS2WS4

WS5

WS3

Biodiversity Service Network

Workflow and experiments to study the ecological niche of the Horseshoe crab (Limulus polyphemus)

Workflows, pipelines and other applications are built from “services”

• Workflows allow to run studies and experiments to process vast amounts of data, repeatedly– Select and apply successive “services”

(data analysis and processing steps)– Import data from own research and/or

from existing public sources– Choose input parameters

• Service nesting– Services can themselves be made up

of combinations of other services or of workflow fragments

A grouping of Web services having related functionality is called a ‘Service Set’

Taxonomy Metagenomics and metagenetics

Ecological niche and population modelling

Ecosystem functioning and valuation

Mapping, visualization, transformation

Catalogue of Life name lookup

QIIME ENM (openModeller)

Get meteor-ological data

Spatio-temporal visualization

GBIF occurrence data retrieval

BOLD PopBio Weather to Biome-BGC data

GeoServer WMS/WFS/WCS

GBIF ChecklistBank

BlastX Biome-BGCmonte carlo

Raster Diff

WoRMS aphia name

Sequence (OTU) clustering

Biome-BGC sensitivity anal.

ISO Country Code

PESI name Functional diversity

Data-Model harmonization

DwC-A to JSON shim

Checklist Cross-mapping

Taxonomic diversity

Biome-BGC CARBON

DwC-A to CSV shim

?

Taxonomy &Systematics

Ecological niche andpopulation modelling

Ecosystem functioningand valuation

?

Genes-Species-Specimens(multi-scale linkages)

Citizen Science &Observations

Mapping, visualization andtransformation services

Service sets driven by science and policy needs

• CO2 emissions continuously increasing– 10 GtC in 2010; Sequestration is the sustainable

process to mitigate the effects

• Over the past 50 years, humans have changed ecosystems– resulting in a substantial and largely irreversible loss

of biodiversity

• Invasions of alien species– A leading cause of biodiversity loss and related

economic damages. They degrade ecosystem services, generate human health problems and impact outdoor recreation.

“transportation with ships is a high risk to

spread the species to these spots”

Stelzer et al 2013

Source: NOAA

Service sets driven by science and policy needs

• CO2 emissions continuously increasing– 10 GtC in 2010; Sequestration is the sustainable

process to mitigate the effects

• Over the past 50 years, humans have changed ecosystems– resulting in a substantial and largely irreversible loss

of biodiversity

• Invasions of alien species– A leading cause of biodiversity loss and related

economic damages. They degrade ecosystem services, generate human health problems and impact outdoor recreation.

“transportation with ships is a high risk to

spread the species to these spots”

Stelzer et al 2013

Source: NOAA

Modellingecosystem services

ModellingCO2 sequestration

Calculating measures of genetic diversity

Assessing adaptationto changing conditions

Supporting processesof conservation

Assisting invasivespecies management

• Connecting biology and IT communities– Distinct languages, different understandings– Service Network approach connects them

• Supporting use cases we know today ...– … and use cases in the future that we cannot

yet imagine

• Different Service Providers are good (competent) at different things

• Deals with multiple jurisdictions and supports a business model– Leading to sustainability

Why do we need them?

Scientists’ perspectives

Info

rmati

on T

echn

olog

ists

’pe

rspe

ctive

s

Biodiversity studies & experiments

Services for biodiversity science

compose to support

ICT Technical Capabilities

ICT Technical Elements

combine to deliver

combine to support

How to build Service Networks and Service Sets?Discipline

Scientists

Scientific PAL

Technical PAL

Scientific and Technical Service Providers

ScientificRequirements

Translation

TechnicalRequirements

TechnicalCapabilities

ScientificCapabilities

ApplicationServices Team

Prioritisation

Support Centre

Training &Issue Resolution

Service LevelRequirements

Sustainability

Community

CommunitySource: M.Obst

Objectives in building Service Networks

1. Services MUST be secure, scalable, reliable, and well-documented

2. Services CAN be deployed on standard cloud configurations

3. Services SHOULD be implemented according to recognised ‘best practices’

4. Services MUST be discoverable

Services MUST be secure, scalable, reliable, and well-documented

Users’ workflows and applications

Sustained Service and Data ProvidersGBIF, CoL, ITIS, OBIS, WoRMS,EBI, BGBM, CRIA, EoL, BHL, ALA, etc. + many many more

Recognised and stable Resource ProvidersNational, EGI.eu, PRACE, commercial, etc.

Services CAN be deployed on standard cloud configurations

• Deployment on standard cloud platforms for better availability and scalability

– 2-tier trust relations

• Pilot service for ‘e-Infrastructure Commons’– Private / public collaboration– Builds in 2014

?

Services SHOULD be implemented according to recognised ‘best practices’

• Best practices for improving ease of use and scalability– “The perfect API” : What would be its characteristics?– Wider adoption of standard data and parameter formats to promote

interoperability

• Classes of “shims”– Utility, Format handling, Data from sources

• Founded on standard infrastructure– Use of standard approaches

• to e.g., authentication, authorization– For data storage and staging– For persistent identifiers– For metadata generation– Etc.

?

Services MUST be discoverable

www.biodiversitycatalogue.orgA fully curated, well-founded catalogue of

Web services for biodiversity science

• GEO BON– Biodiversity Observation

Network– By 2015

• Functional infrastructure“from observations to Essential

Biodiversity Variables (EBV) and derived indicators”

• Using a plug-and-play, service-oriented approach

• Coordinated through a registry system

• Linked to the GEOSS Common Infrastructure

Where’s this all going?

• ALA• SiBBr• SpeciesLink• GBoWS• DataONE• GBIF• GEOSS• LifeWatch• SANBI

Towards interoperability guidelines for an Integrated Virtual Environment (IVE) for Biodiversity Science at the

international level

What’s trending?

• Software is a key enabling technology– In a distributed and mobile world, this means Web apps

• Web apps means APIs. APIs give access to services– How to create connections between APIs that don’t usually

talk to each other?

• WebApp automation like:– IFTTT – “IF this THEN that”– Zapier

• API / service management & monetization like:– 3Scale

Take home messages

• Separate:– Services from underlying infrastructure– Applications, workflows and VREs from the services

• Ensure Service Networks are built using standard Web 2.0 technologies

• BiodiversityCatalogue is the well-founded place to register and discover services

• There’s much more to do!