Difference between revisions of "SWAO:Conference Call 20210607"
Arwesterinen (talk | contribs) (→Workshop and Communique on Interoperability for Situation Management) |
Arwesterinen (talk | contribs) (→Mapping Events (from Narratives)) |
||
| Line 63: | Line 63: | ||
=== Mapping Events (from Narratives) === | === Mapping Events (from Narratives) === | ||
| − | AndreaW overviewed [https://swao.s3.amazonaws.com/Mapping+Event+Text+to+RDF.pptx | + | AndreaW overviewed [https://swao.s3.amazonaws.com/Mapping+Event+Text+to+RDF.pptx her presentation]. There was some confusion over what was being discussed. So, the name of the PowerPoint was updated to "Creating a Knowledge Graph/RDF Triples from Events in Text". |
| − | The presentation shows parsing of text from a narrative from the US Holocaust Museum. The parse was done using [https://spacy.io/ | + | The presentation shows parsing of text from a narrative from the US Holocaust Museum. The parse was done using [https://spacy.io/ spaCy]. The analysis used basic information such as parts-of-speech, dependency parsing, definitions and synonyms to tie the text back to an ontology of events (i.e., [https://github.com/ontoinsights/deep_narrative_analysis/tree/master/ontologies this ontology] in GitHub). |
An outstanding question was what was the spaCy training data. That was answered post-call via email. The spaCy transformer model (used in this work) was trained on the OntoNotes 5 corpus. | An outstanding question was what was the spaCy training data. That was answered post-call via email. The spaCy transformer model (used in this work) was trained on the OntoNotes 5 corpus. | ||
| − | There was also a brief sidebar discussion of KenB's [https://ken.baclawski.com/pub/2018/21/public.pdf | + | There was also a brief sidebar discussion of KenB's [https://ken.baclawski.com/pub/2018/21/public.pdf Observer Effect paper], and how observations are captured in the Event ontology. AndreaW referenced the event-subclasses.pdf image (in the GitHub repository linked above) to show that observations (Measurements) are captured as Events with ReportedValues, at a time and place. And, each Measurement instance can have a topic - which is the thing measured/observed/reported. |
=== Ontology Articulation Guidelines === | === Ontology Articulation Guidelines === | ||
Revision as of 03:34, 9 June 2021
| Number | 91 |
|---|---|
| Duration | 1 hour60 minute 3,600 second 0.0417 day |
| Date/Time | June 07 2021 18:00 GMT |
| 11:00am PDT/2:00pm EDT | |
| 7:00pm BST/8:00pm CEST | |
| Convener | Andrea Westerinen |
IAOA Semantic Web Applied Ontology (SWAO) SIG
THIS MEETING WILL TAKE PLACE ON MONDAY 7 JUNE
Meetings are normally on the first Monday of the month at 2pm Eastern Time.
- We use Zoom for these meetings - details at Connection Details
Contents
Agenda
- Latest post for the IAOA website available on main IAOA page, https://iaoa.org/
- Note that the new convener was not announced as it is mainly an administrative thing
- Actions from last meeting
- AndreaW and MikeB input to RobertR on his paper
- AndreaW details on mapping events to KG (overview to be discussed below)
- RobertR new foundational ontology?
- Status of discussions at workshop and communique on "Interoperability for Situation Management and Decision Making" (KenB)
- Is there a new draft?
- Mapping events to KG (AndreaW)
- Ontology Articulation Guidelines (AndreaW)
- Proposed outline for several papers (AndreaW, to be available for discussion on the call)
- Housekeeping
- AoB
- Next Meeting? (July 5th is US holiday)
Proceedings
Pre-call discussion
KenB provided a short introduction to a new query language that he is developing - KGQL (the Knowledge Graph Query Language). It will be discussed in an upcoming paper in a special issue of the IAOA Applied Ontologies journal. (More on this at the end of the meeting notes.)
Actions Review
- AndreaW provided input to RobertR on his paper dealing with concepts/terms used in ontologies (esp top-level ones); MikeB was not yet in attendance and could not report CLOSED
- Presentation on mapping events later in today's meeting CLOSED
- RobertR was not in attendance - new foundational ontology action still OPEN
Workshop and Communique on Interoperability for Situation Management
KenB reported that the second workshop at the IEEE International Conference on Cognitive Methods in Situation Awareness and Decision Support (CogSIMA) happened in May, and that there were many great discussions. He has not yet had time to consolidate the feedback into a new release of the communique. KenB previously shared the draft (as discussed in the Conference_Call_20210510 notes).
NEW ACTION: Check on availability of a new draft of the Interoperability Communique
Mapping Events (from Narratives)
AndreaW overviewed her presentation. There was some confusion over what was being discussed. So, the name of the PowerPoint was updated to "Creating a Knowledge Graph/RDF Triples from Events in Text".
The presentation shows parsing of text from a narrative from the US Holocaust Museum. The parse was done using spaCy. The analysis used basic information such as parts-of-speech, dependency parsing, definitions and synonyms to tie the text back to an ontology of events (i.e., this ontology in GitHub).
An outstanding question was what was the spaCy training data. That was answered post-call via email. The spaCy transformer model (used in this work) was trained on the OntoNotes 5 corpus.
There was also a brief sidebar discussion of KenB's Observer Effect paper, and how observations are captured in the Event ontology. AndreaW referenced the event-subclasses.pdf image (in the GitHub repository linked above) to show that observations (Measurements) are captured as Events with ReportedValues, at a time and place. And, each Measurement instance can have a topic - which is the thing measured/observed/reported.
Ontology Articulation Guidelines
The latter half of the call was spent reviewing the Articulation Guidelines document. The text in that document was assembled from discussions over the previous 5 months, regarding what should be included. As the attendees reviewed the document, additions and clarifications were made inline.
The document linked above is the version with changes. It is likely that several papers (or perhaps a book) could result from the work. It was decided that any articulation of guidelines mandated first explaining the word, "ontology".
NEW ACTION: AndreaW to take a stab at initial text for "What is an ontology?", for review in the next meeting.
AOB
KenB mentioned that the call for papers for a Special Issue of IAOA Applied Ontology will be announced. That issue will be based on the 2020 Summit (Knowledge Graphs). There was a short discussion of timing for Abstracts and final submissions (~Sept 1).
Next Meeting
It was decided that the next SWAO SIG meeting would be on July 12.
Attendees
- Ken Baclawski
- Andrea Westerinen
- Todd Schneider (conflict starting at 2:30PM)
- Mike Bennett
Video Recording
Available [1].
