View source for AI Use Case Scoping: Discussion Resources

[[File:ocfcLogo.png|200px]]

>> [[OCF_Collab_Network|OCFC Home]] >> ''Scoping AI Use Cases''

== Scoping AI Use Cases for Scaling Competency Data ==
The AOC (formerly RWSC) interspersed AI scoping in multiple meetings and held two meetings focused on use cases for scaling competency data.  The following resources were used for discussions.
*[https://www.uschamberfoundation.org/sites/default/files/T3%20Report_Skill%20Competency%20Data%20Translation%20and%20Analysis_Dec2020_FINAL.pdf T3 Phase 2 landscape analysis for skill and competency data] published in 2020
**[https://docs.google.com/document/d/1pVpr-irzRY4DIVxDdfjgigJMZVZmidCKqUTt5MiJIy4/edit AI Use Cases] pulled directly from the Phase 2 landscape analysis for skill and competency data report
*[https://docs.google.com/document/d/1Iebix9yQqMm7QY49nAuTOVnyZ1HSL0ZFex_KoF3uijQ/edit AI Discussion Guiding Questions]

== Themes From Discussions ==
AI will play a vital role in scaling competency data that is important to all T3 Networks.  Each T3 network need reliable sources for competency data for numerous use cases and actors.

'''''Scaling Competency Alignments and Mapping'''''

* generating semantically-based competency alignments between competencies and other resources (e.g., credentials, assessments, learning opportunities, jobs, work roles, tasks, etc..)

'''''Scaling Competency Analysis'''''

* accelerating currency of competency trends
* drilling down to micro or up to macro details such as geographic significance
* protecting identity (companies and individuals)
* providing measures of competency alignment
* removing bias and increasing equity
*validating data-justified specificity of statements, e.g. determining if competencies within a task are defined at a measurable level or left to too much human interpretation and if sufficiently contextualized. I.e. finding the poorly defined statements.

'''''Scaling Competency Data Search Sources'''''

* extracting competencies with context from documents (e.g., competencies from job postings and job descriptions with the context of the job specialty)
* improving cross-competency searching (e.g., automating alignment for competency disambiguation)
* living representations of competencies (rather than static one-time information)
* providing curated data for research and benchmarking

== Additional Resources ==
*May 4, 2022 RWSC Meeting: Guest Presentation on skillsync AI services by Robby Robson, Elaine Kelsey, and Elliot Robson with Eduworks 
**[https://docs.google.com/presentation/d/1m8NWB9qEbu4oB8ZOfakmQE_5R8Gh6tW6DbifcZ9iVrg/edit?usp=sharing Presentation] (slides 9-12) and starts at about 8:18 mins in the [https://drive.google.com/file/d/1EKVkvbW1KET1wkU-ckeju1nHYe2xImVj/view?usp=sharing Video]
***skillsync Resources:
****[https://ojs.aaai.org/index.php/aimagazine/issue/view/490 AI Magazine Spring Issue Vol. 43, Issue 1]:  [https://drive.google.com/file/d/1dtaCSihPajLMNawlK7-TivSkESOHXH4z/view Intelligent Links: AI-Supported Connections between Employers and Colleges] (pages 75-82 )
****[https://evolllution.com/revenue-streams/workforce_development/supporting-companies-and-colleges-as-they-reskill-the-workforce-of-the-future/ Supporting Companies and Colleges As They Reskill the Workforce of the Future], October 18, 2021
****[https://skillsync.com/ SkillSync Website]
*April 20, 2022 RWSC Meeting: Guest facilitation by Matt Gee, Co-Founder & CEO at BrightHive / Senior Researcher at the University of Chicago
**Two themes came from the discussion: (1) improving cross-competency framework search (automating alignment) and (2) providing curated data for research and benchmarking.
**[https://docs.google.com/presentation/d/1u99nJ3sppJ1X6pXoah0hZPdM9OzexuPWv0pnjxA3Gh0/edit#slide=id.g124f0de3d0c_0_29 Presentation] (slides 8-12), [https://drive.google.com/file/d/1B3TncLdJ1jvkpETL8Vz-G0CvzUyZFPBV/view?usp=sharing Video], [https://drive.google.com/file/d/1S2j3zzuYa5k7uEL8FjhjaV5QC-s0lGBG/view?usp=sharing Audio]
*There is some similar work happing in Europe in projects related to the European Skills, Competences, Qualifications and Occupations Initiative (ESCO) that reinforces the message about how AI and machine learning can help in building frameworks of competences around which education and employment related services can cohere.  Here are three reports from ESCO that are worth a look, one relatively recent and narrow, the other two a year old and broader (all PDFs):
**[https://esco.ec.europa.eu/system/files/2022-03/theRoleOfContextualInformationWhenConnectingDataToTheESCOOccupationsPillarUsingArtificialIntelligence_1.pdf The role of contextual information when connecting data to the ESCO Occupations Pillar using Artificial Intelligence]
**[https://esco.ec.europa.eu/system/files/2022-03/leveragingArtificialIntelligenceToMaintainTheESCOOccupationsPillarReport%20%281%29.pdf Leveraging Artificial Intelligence To Maintain The ESCO Occupations Pillar]
**[https://esco.ec.europa.eu/system/files/2022-03/leveragingArtificialIntelligenceToUpdateTheESCOOccupationsPillarReport%20%281%29.pdf Leveraging Artificial Intelligence To Update The ESCO Occupations Pillar]
**[https://drive.google.com/file/d/1qezXVnJ0IQay0fqALwMzM7FH_vRixQSX/view?usp=sharing Slide deck] from a presentation about this work given a couple of months ago at the ESCO 1.1 launch, for which there is a video at <nowiki>https://www.youtube.com/watch?v=YPECXVRagu8</nowiki> 
**