Ontolog Forum
| Session | Synthesis |
|---|---|
| Duration | 1 hour |
| Date/Time | 04 May 2022 16:00 GMT |
| 9:00am PDT/12:00pm EDT | |
| 5:00pm BST/6:00pm CEST | |
| Convener | Ravi Sharma |
| Track | Environment Disasters |
Ontology Summit 2022 Synthesis
Dealing with Disasters
The COVID-19 pandemic as well as other pandemics and disasters have prompted an impressive, worldwide response by governments, industry, and the academic community. Ontologies can play a significant role in search, data description, interoperability and harmonization of the increasingly large data sources that are relevant to disasters such as the COVID-19 pandemic. The Ontology Summit 2022 examined the overall landscape of disasters and related ontologies. A framework consisting of a set of dimensions was developed to characterize this landscape. The framework was applied to health-related disasters, environmental disasters, as well as aerospace and cyberspace disasters. It was found that there are many cross-domain linkages between different kinds of disasters and that ontologies developed for one kind of disaster can be repurposed for other kinds. A representative sample of projects that have been developing and using ontologies for disaster monitoring and response management is presented to illustrate best practices and lessons learned. The Communiqué ends by presenting the findings and recommendations of the summit.
Agenda
- Synthesis
- Dr Ravi Sharma - Earth Observations to monitor disasters - sensors, platforms, image-data and metadata (15-20 minutes) Slides Video Recording YouTube Video
- Synthesis of the Environmental Disaster Track. Draft Synthesis Video Recording
Conference Call Information
- Date: Wednesday, 04 May 2022
- Start Time: 9:00am PDT / 12:00pm EDT / 6:00pm CEST / 5:00pm BST / 1600 UTC
- ref: World Clock
- Expected Call Duration: 1 hour
- The Video Conference URL is https://bit.ly/3rTKSGQ
- Meeting ID: 881 4427 2329
- Passcode: 553714
- Chat Room: https://bit.ly/37g93pC
- If the chat room is not available, then use the Zoom chat room.
- One tap mobile
- +13126266799,,88144272329#,,,,*553714# US (Chicago)
- +16465588656,,88144272329#,,,,*553714# US (New York)
- Find your local number: https://us02web.zoom.us/u/k7asjiSSZ
Attendees
- Alex Shkotin
- Andrew Dougherty
- Gary Berg-Cross
- Janet Singer
- Ken Baclawski
- Marcia Zeng
- Michael Singer
- Ram D. Sriram
- Ravi Sharma
- Robert Rovetto
Discussion
[11:32] Ravi Sharma: Welcome everyone to this track 3 synthesis session.
[12:41] Gary Berg-Cross: For the synthesis an issue I had with Dave Jones and Karen Moe's (ESIP) presentation is that it shows how far from semantics many trusted data efforts are. More specifically How do we enhance semantics of the ORLs they talked about? What is Metadata completeness that people filled out? It means what? Also in their presentation it is unclear what automated intelligence means.
[12:46] Gary Berg-Cross: Jones and Moe their presentations all provide an organizational view of data and operation on disasters. It may remain a challenge to get some knowledge-level processes, such as data harmonization, engaged in these organizations and their activities.
[12:48] Gary Berg-Cross: In Jano presentation "Can We Make Useful Data Re-usable?" we get a bit more semantics thru KnowWhereGraphs and the geoEnrichment approach.
- We hear a bit more about the power of environmental intelligence applications
- We hear about Issue With Portals which just reach isolated data sets.
[12:52] Gary Berg-Cross: A key point about KGs from Jano is that we are still in the early stages of this work - we know almost nothing about them and the associated Knowledge Graph Ecosystem.
[12:52] Gary Berg-Cross: Jano's work also goes beyond classical KR to include ML embeddings.
[12:56] Gary Berg-Cross: The weAdapt effort falls between the Portal efforts and Jano's KG their Connectivity Hubis designed to link people with the gross knowledge they need. It is better than a simple portal it has some taxonomy.
[12:56] Gary Berg-Cross: weADAPT uses keyword tagging technology to link related content together and make it easier to explore and discover content. Using a shared, well-described taxonomy to annotate content on websites is a key step for realizing Linked Data.
[12:57] Gary Berg-Cross: Some background on weAdapt taxonomy at https://www.weadapt.org/knowledge-base/climate-knowledge-brokers/transforming-knowledge-management-for-climate-action
[12:59] Gary Berg-Cross: Jano makes several points we can leverage of going Beyond Portals to what KGs provide:
- We don't serve links to datasets; we serve every single data record in these datasets in a searchable and AI-ready way.
- We don't provide yet another region (e.g., county) specific portal; we serve data across many different types of region identifiers.
- We don't make you download, overlay, and analyze different datasets on various topics relevant to your information needs; we answer your questions directly and based on pre-integrated data.
- We don't go online with the next technology change, e.g., Flash, Silverlight; we rely on a distributed, internationally standardized open-source technology paradigm.
[13:00] Gary Berg-Cross: Also for our communique Jano includes Key Research Questions we might discuss:
- Spatiotemporally explicit KG embeddings that are invariant under syntactic changes
- How to Learn geographic knowledge graph summaries for apportioned geoenrichment
- Quantifying and representing regional differences in the conceptualization of geographic space
- How to detect and mitigate bias in geographic knowledge graphs
Resources
- Earth Observations to monitor disasters - sensors, platforms, image-data and metadata
- Draft Synthesis of the Environmental Disaster Track