Hackpads are smart collaborative documents. .

Stephan Zednik

460 days ago
leeb5@rpi.edu eScience June 16, 2016
 
Patrick W
  • General Information
 
Call In
Marshall M 1.  Please join my meeting.
2.  Use your microphone and speakers (VoIP) - a headset is recommended.  Or, call in using your telephone.
Dial +1 (408) 650-3123
Access Code: 333-351-829
Audio PIN: Shown after joining the meeting 
Meeting ID: 333-351-829
 
leeb5@rpi.edu
  • Attendance
Jim M Jim McCusker
Han W Han Wang
leeb5@rpi.edu Marshall Ma
Hao Z Hao Zhong
Peter F Peter Fox (in office working on a proposal)
 
  • Notes
  • Producer vs Consumer on Data Quality
  • Quality control vs quality assessment
  • Control
  • Reflects "happiness" of the data
  • Whether the algorithms converge and so forth
  • Assessment is an analysis "after the fact" 
  • Biased by point of view
  • Facets of Quality
  • Accuracy
  • Completeness
  • Consistency
  • Resolution
  • Ease of Use
  • Latency
  • Have projects that focus on different level of quality
  • What other factors when looking at Facets of Quality?
  • Factors named before lean towards NASA data
  • What factors do we think exist that are domain independent?
  • List of steps to follow independent of domain that will allow us to analyze data quality
  • General method to extend the currently existing facets to begin to apply it to other domains
 
  • Discussion
  • Most current facets are dataset dependent
  • Cannot neccesarily use this generally for projects that use multiple datasets together
  • NASA data tends to be large, does this method still hold for smaller datasets?
  • If it has one and not the other, is the data quality still good?
  • Example: Museum Curator
  • What data quality facet is most important to you give a use case to put wildfires together
  • Current methodology: bundle the metadata with the data so that consumers can make the decision on quality
  • Marshall: Vocabulary quality developed independently and they also have Completeness, Consistency, and Accuracy
  • Spatial and Temporal commonly show up again within the main topics
  • Peter and Stephan: Completeness used to be Coverage, but it usually implies spatial, but not temporal so Completeness shifts towards including both.
  • Genetics, usually don't need spatial, but maybe temporal.  Quality may need to be separate from utility
  • Are the answers there or not?  Finding the quality of the data.
  • Use Case?  Mass Spectroscopy(sp?) - drift over a set of measurements
 
 
 
474 days ago
Congrui L Congrui Li
Hao Z Hao Zhong
Stephan Z Stephan Zednik
  •  
 
Topic: Insider Threat Ontology
 
  • Background
  • Malicious Insider - internal employee that may perform malicious actions
  • Ontology abstracts the idea of threats consistently across organizations
  • Trades expressiveness for inter-operability
  • Implementation
  • Distinguish between Actions (observable) and Events (inferrable)
  • Validation
  • "The insider stole a co-worker's password credentials to log into the system and commit fraud."
  • Start with Natural Language, then use this model to translate into a Semantic structure
 
Discussion
 
  • Difference between Asset and Information?  (Credit Card Asset vs Credit Card Information)
*Money stolen from a credit card is the asset, the card number is the information
  • Authors created semantic model in last slide manually
*Validating the ontology is expressive enough
*Possible Research: Automate NLP to generate these Semantic models
*Linkepedia?  Perhaps a more domain specific domain ontology
*Owl File not available, but forth
 
481 days ago
Deborah M Discussion Topics
  • best practices google doc <get pointer>
  • just navigating around our github content related to this
Han W
  • where are our ontologies?  and related sparql endpoints
  • where are our ontologies?  and related sparql endpoints
 
 
Marshall M 20160526 eScience Meeting 
 
Patrick W
  • General Information
  • This Meeting
  • Previous Meeting
  • Next Meeting
 
Call In
Marshall M 1.  Please join my meeting.
2.  Use your microphone and speakers (VoIP) - a headset is recommended.  Or, call in using your telephone.
Dial +1 (408) 650-3123
Access Code: 333-351-829
Audio PIN: Shown after joining the meeting 
Meeting ID: 333-351-829
 
Patrick W Attendance
Marshall M Marshall Ma
Han W Han Wang
Ian Gross
Jim M Jim McCusker
anirudh p Hao Zhong
Deborah M Zhen Liu 
Sabita Acharya  (will be remote after this week)
Liyu Pan <panl2@rpi.edu>
Rui Yan
 
 
Regrets:
  • Henrique O. Santos <hensantos@gmail.com>   
  • Paulo Pinheiro at RPI <pinhep@rpi.edu>
  • katie chastain (already on list)
Peter F
  • Ahmed (on travel, on list)
  • Sophie Kolankowski (getting married!)
 
Marshall M
  • Propose and discuss topics and questions for the coming eScience meetings
 
anirudh p Location - Place, Event, geographical etc
Ontology Change Template - May not apply for some ontology
Minimum subset of the template that needs to be filled. 
 
Patrick W Action Items
anirudh p
  • Ontology evolution talk based on Linyun's Work
  • Semantics of Quality and Bias Representations
  • Location representations in PROV 
  • Insider's Threat - Rui
 
Deborah M Please add me to the list  (the first 5 people below are all working with deborah this summer)
Patrick W
  • rashidshabbir@gmail.com
anirudh p
  • grossi2@rpi.edu
  • sachar4@uic.edu
Deborah M
  • Zhen Liu <zl1471@nyu.edu>
  • Liyu Pan <panl2@rpi.edu>
 
Patrick W Potential Topics
  • ontology best practices
  • Github Repositories
  • ontology versioning
  • ontology annotation
  • content negotiation and discoverability
Deborah M
  • registering (and advertising) our ontologies
Patrick W
  • easily creating semantic content (ontology editors as well as instance editors)
Peter F
  • Location representations (Marshall and Peter from Vespucci last June)
Marshall M
  • can also relate to work of the 'Spatial Data on the Web' OGC-W3C joint WG
Peter F
  • Location in PROV-O (Marshall and Peter)
Stephan Z
  • ?is this related to the location representations bullet above? Yes.
  • W3C PROV definition of Location: A location can be an identifiable geographic place (ISO 19112), but it can also be a non-geographic place such as a directory, row, or column. As such, there are numerous ways in which location can be expressed, such as by a coordinate, address, landmark, and so forth.
Deborah M
  • Ontology evolution and curation among Big / Broad communities
...
918 days ago
Unfiled. Edited by Stephan Zednik 918 days ago
Stephan Z W3C Recommendation - Linked Data Platform
 
 
Documentation
Live Owl Documentation Environment (LODE) is a service that will generate HTML documentation pages for OWL ontologies based on the ontology classes/properties and their annotations.
  • supports many owl, rdfs, and dc annotation properties
  • encourages use of annotation properties
 
This is very useful to create web-accessible human-readable documentation of the ontology.  Use of LODE is highly recommended.
 
 
876 days ago
  • Next Meeting
 
Attendance
Patrick W Patrick
Linyun F Linyun
Stephan Z Stephan
 
  • Topics of Interests
  • Relevant Papers:
 
  • Agenda
  • Discussion (15-20 min max)
  • Summary (3-5 min max)
 
Action Items from Last Time
 
  • Notes
  • The relationship between verisoning and the rest of the evolution cycle elements
Linyun F
  • Versioning as a way to encode and manage changes (i.e., the rest of the evolution cycle elements/activities)?
  • approaches/languages to quantify/summarize/encode changes
  • GCIS ontology change as a concrete example
 
Action Items
  • Anirudh to jot down ontology changes from GCIS second phase
 
932 days ago
zhongh3@rpi.edu eScience Meeting 20150302.1100 Mon
 
Patrick W
  • General Information
Patrick W
  • Next Meeting
 
zhongh3@rpi.edu Please join my meeting, Mar 2, 2015 at 11:00 AM EST.
 
Dates and Speakers
Date Weekday Time Speaker Topic
Feb 02 Monday 11am ET Linyun Fu Ontology Interest Group
Feb 16 Monday 11am ET Stephan Zednik Vocabulary Services
Mar 02 Monday 11am ET NA Catchups & Discussion
Mar 16 Monday 11am ET Linyun Fu Linked Vocabulary Schema
Mar 30 Monday 11am ET
Apr 13 Monday 11am ET
Apr 27 Monday 11am ET
May 11 Monday 11am ET
Possible Speakers and Topics for the first week:
 
 
Patrick W Call In
 
Attendance
Stephan Z
  • Stephan Zednik (remote)
  • Peter
Patrick W
  • Patrick (remote)
 
 
  • Around the room (5-10 minutes max) 
  • Anything following up from last week (active items)
  •  
  • Talk
  • Discussion (15-20 min max)
  • Summary and What's Next (3-5 min max)
 
Action Items from Last Time
Stephan Z
  • Stephan - find old proposal for update to PML and share with Patrick as example for proposal to update ontology
  • Stephan - write down previous ontology publication practices
Patrick W
  • Patrick - proposal for TW ontology change
  • Patrick - proposal for TW Ontology Best Practices working group
Linyun F
  • Linyun - initiate a mail thread on the [ontology] list on specializing the LDA ontology   (please expand LDA)
  • a mixture of rationale and implementation
 
 
Stephan Z
  • Review previous meeting's action items
Patrick W
  • DOAP (Description of a Project) - Software as a project
 
Action Items
Deborah M
Linyun F
  • Linyun Fu: Prepare a talk about LVA ontology design in two weeks.
  • make sure not to ignore Epimorphics' credit.
Patrick W
  • Use rpi handle system to generate handles for all of our instances
  • Would love to see someone look into purl and give a presentation!
 
 
 
918 days ago
Mar 16 Monday 11am ET Linyun Fu Linked Vocabulary API
 
  • Stephan (remote)
Linyun F
  • Linyun
anirudh p
  • Anirudh
Patrick W
  • Patrick
 
Agenda
  • Action Items from Last Time (10-15 minutes max) 
 
  • Linyun Fu: Prepare a talk about LVA ontology design in two weeks.
  • make sure not to ignore Epimorphics' credit.
  • Patrick: NOT YET - Use rpi handle system to generate handles for all of our instances
  • Patrick: More of a suggestion - Would love to see someone look into purl and give a presentation!
 
  •  
 

Contact Support



Please check out our How-to Guide and FAQ first to see if your question is already answered! :)

If you have a feature request, please add it to this pad. Thanks!


Log in