From OntologPSMW

Jump to: navigation, search
[ ]
    (1)
Session Anirudh Prabhu
Duration 1 hour
Date/Time 26 February 2020 16:00 GMT
9:00am PDT/12:00pm EDT
5:00pm BST/6:00pm CEST
Convener Gary Berg-Cross
Track How

Contents

Knowledge graphs, closely related to ontologies and semantic networks, have emerged in the last few years to be an important semantic technology and research area. As structured representations of semantic knowledge that are stored in a graph, KGs are lightweight versions of semantic networks that scale to massive datasets such as the entire World Wide Web. Industry has devoted a great deal of effort to the development of knowledge graphs, and they are now critical to the functions of intelligent virtual assistants such as Siri and Alexa. Some of the research communities where KGs are relevant are Ontologies, Big Data, Linked Data, Open Knowledge Network, Artificial Intelligence, Deep Learning, and many others.     (2A)

Agenda     (2B)

Abstract: Knowledge graphs have been used with reasoners to make inferences about the data, based on the assertions and axioms that have been written. But a knowledge graph, which houses complex, multi-dimensional data also contains hidden patterns and trends which cannot be explored simply by using reasoners. In this talk, we present various approaches for using data extracted from knowledge graphs to gain scientific insights.     (2B2)

Bio: Anirudh Prabhu is a PhD student at Rensselaer Polytechnic Institute, working under Prof. Peter Fox. Research interests include Data Modeling, Semantic E-science, and Data Visualization. Current research work focuses on applying Data Science and Machine Learning to solve Earth and Environmental Science problems. For more information see Anirudh Prabhu Website     (2B3)

Conference Call Information     (2C)

Attendees     (2D)

Discussion     (2E)

[12:11] RaviSharma: Anirudh believe you are using Sparql and Allegrograph?     (2E1)

[12:17] RaviSharma: do all data have to be formed as triples or your software and algorithms discover the relations / links?     (2E2)

[12:19] BobbinTeegarden: Where are the slides for download?     (2E3)

[12:21] KenBaclawski: The slides are available at https://go.aws/3a9Niax     (2E4)

[12:22] ToddSchneider: What were the sources of data used?     (2E5)

[12:23] John Sowa: The application is interesting, but it could be developed with information in many different formats.     (2E6)

[12:24] John Sowa: What features of KGs are more/less useful than other representations?     (2E7)

[12:31] ToddSchneider: What definition of 'Knowledge Graph' was, or is, used?     (2E8)

[12:33] RaviSharma: using these you can predict natural and industrial impact based mineral distributions over time     (2E9)

[12:34] janet singer: @John In the 1970s John Warfield began speaking how one could use the model exchange isomorphism to go between logical, graphic (network) and matrix representations in group modeling efforts. Each had advantages that could be leveraged     (2E10)

[12:34] RaviSharma: how do you distinguish between link and nodes?     (2E11)

[12:37] janet singer: So this is interesting not because it is new, but specifically because it represents a maturing of that core idea.     (2E12)

[12:37] ToddSchneider: What or which ontologies?     (2E13)

[12:38] Ram D. Sriram: @Janet: Good point     (2E14)

[12:39] Ram D. Sriram: @John: Good question. I think one of the key issues now is to standardize the terminology, construction, and visualization of KGs     (2E15)

[12:41] janet singer: An early paper by Warfield http://bit.ly/32BlQ2U     (2E16)

[12:43] BobbinTeegarden: The great part here is being able to SEE the patterns, and discover patterns from different perspectives. It's the 'shape' of the 'sub-ontologies' in the knowledge graph that give insights...     (2E17)

[12:47] RaviSharma: Anirudh - great presentation and techniques - please let us know when you define the general case relating to nodes links and ontologies.     (2E18)

[12:49] Ram D. Sriram: @john: I remember you had a visualization framework in your DARPA project     (2E19)

[12:51] BobbinTeegarden: @JohnSowa Where are all the visualizations of an ontology from past? We did molecular modeling etc but I don't think they were interactively explorable     (2E20)

[12:53] RaviSharma: janet yes +1     (2E21)

[12:53] David Eddy: I duly note (for me news) it took about 200 years for knowledge to progress from alchemy to the periodic table.     (2E22)

[12:54] Chris A: Hi. Very interesting. You have 2 jupyter notebook links on slide 28, however, when clicking on them they take me to a password protected page. Are we able to gain access to them?     (2E24)

[12:54] TerryLongstreth: @DavidEddy: but now we have computers!     (2E25)

[12:54] janet singer: @David: Yes, definitely     (2E26)

[12:55] RaviSharma: janet - we need to create a summary of what John is saying and also to see how to relate it to speakers' works     (2E27)

[12:56] Gary: John is good at providing summaries of what he has said.     (2E28)

[12:57] BobbinTeegarden: The interesting part here is the visualization (goes directly to the 'right brain') and can give wetware experts Aha experiences...     (2E29)

[12:57] RaviSharma: Janet - we are seeing practical and new ways of depicting information as today which is of value     (2E30)

[12:58] RaviSharma: I agree with Janet and Gary's comments.     (2E31)

[13:00] RaviSharma: Bobbin - yes     (2E32)

[13:02] David Eddy: @Terry... surely you're not arguing they've helped? Computers have just made the problem datasets hugely bigger.     (2E33)

[13:02] BobbinTeegarden: What is Sweet? Where find it?     (2E34)

[13:02] RaviSharma: Anirudh - what do you do to align meanings in different domains     (2E35)

[13:04] Gary: Semantic Web for Earth and Environment Technology Ontology it is on the Bio Ontology Portal     (2E36)

[13:05] David Eddy: As John Sowa pointed out recently... the semantics got dropped from the SemanticStack     (2E37)

[13:07] janet singer: @John @David: The idea of representation-independent conceptual schemas should definitely be central in the communiqué.     (2E38)

[13:07] RaviSharma: john suggests computers to be able to address follow on Qs for automated negotiation or clarification?     (2E39)

[13:07] BobbinTeegarden: URL for Sweet or Suite?     (2E40)

[13:07] Evan Wallace: I am surprised that people are still using SWEET. I thought that NASA had effectively deprecated it; functionally replacing part of it with QUDT.     (2E41)

[13:07] David Eddy: @Janet... humans do not take kindly to unique meaningless numbers.     (2E42)

[13:07] Gary: SWEET     (2E43)

[13:08] TerryLongstreth: ISO TC 37's focus is terminology     (2E44)

[13:08] RaviSharma: John agrees that communities and organizations are working hard on developing terminologies using NL     (2E45)

[13:08] David Eddy: ... and please to include un-natural language     (2E46)

[13:09] Gary: @Evan, no SWEET is being updated actively as part of ESIP.     (2E48)

[13:09] RaviSharma: John - Math is queen of Sciences     (2E49)

[13:09] Evan Wallace: Interesting.     (2E50)

[13:10] janet singer: @John: That negotiation process of starting with natural language is why Warfield developed interpretive structural modeling.     (2E51)

[13:10] Evan Wallace: @Bobbin, that's it.     (2E52)

[13:10] RaviSharma: Yes Evan thanks     (2E53)

[13:10] TerryLongstreth: @Janet: I agree with you about conceptual schema but we need to link it to it's role as a fulcrum between user views and implementation/storage.     (2E54)

[13:10] ToddSchneider: Meeting ends @13:10 EST     (2E55)

[13:11] RaviSharma: Terry and Janet - and it will be valuable for KGs     (2E56)

Resources     (2F)

Previous Meetings     (2G)


Next Meetings     (2H)