/
Skillabi: Curriculum Skill Relationships (Research)

Skillabi: Curriculum Skill Relationships (Research)

 

Created Date

Mar 31, 2023

Target PI

PI-3

Target Release

TBD

Jira Epic

https://economicmodeling.atlassian.net/browse/NLP-1

Document Status

Draft

Epic Owner

@Kara Foley

Stakeholder

Skillabi Product - @Gavin Esser

Data Science / Models - @Jeff Hoffman (Deactivated) @Seth Friman (Deactivated)

Data Science / NLP ML - @Xiang Li @sahar Nesaei (Deactivated)

Engineering Team(s) Involved

Models Skillabi, NLP ML

PART 1

Customer/User Job-to-be-Done or Problem

JTBD 1: As a Skillabi User and academic/career advisor, I want to guide students along their learning experience based on the curriculum skills that will set them up to achieve their career goal.

  • Help my students optimize their course selection so that they complete their program with the curriculum skills needed to succeed in their chosen career path. Help them to gain/boost particular skills in their program to better prepare for a specific occupation.

  • Help my students gain curriculum skills needed to pivot towards a related career goal, by determining the alternative career goals and associated learning path they should pursue based on existing skills gained and remaining. Help them chart a path to gaining skills needed for a different career goal, minimizing time to degree and maximizing relevance of courses to career goal.

JTBD 2: As a Skillabi user and program owner, I want to identify possible new programs (e.g., certificates, microcredentials) to introduce, drawing from existing courses at my institution, that align with in-demand careers. 

Value to Customers & Users

Solving for JTBD 1: This proposes creating career-based learning pathways using an institution’s curriculum skill content, to guide learners through courses that teach the relevant skills needed for a particular occupation goal, and achieve the following jobs-to-be-done for students:

  • Optimizer: As a learner, when I am progressing throughout my academic program I want to know that the courses I’m taking will equip me with the most important skills needed for my future career, so I can be confident that my credential will be with the cost and I will be able to get employed in my field upon completion.

  • Pivoter: As an adult learner, when I decide to shift towards a new career and/or education goal, I want to determine the best future courses to take while building on course skills I’ve already gained, so I can set myself up with the skills that are most essential for my next career step.

Solving for JTBD 2: Additionally, a model that identifies relevant courses based on the relationship between skills would enable institutions to identify new programs to develop (e.g., microcredentials) using a set of existing courses. Institutions could reformat existing courses to give learners the ability to gain a related set of skills within a skill level.

  • For example, using skill relationships data Skillabi might call up the related courses to recommend an institution turn into a Microcredential as they have similar content and with skills learned at similar levels of proficiency.

Value to Lightcast

  • Allows Skillabi to directly support a stated institutional need to guide learners more effectively through their journey, drive stronger retention and persistence, and prepare those learners for careers.

  • Adds new value for Skillabi by helping users to act on the insights gained in the tool, and make connections across courses/programs.

  • Unlocks the ability for a unique Lightcast/Skillabi strength – curriculum skills – to be used in new value-add and revenue-driving ways, including new program creation.

Target User Role/Client/Client Category

This feature is being built for Skillabi users – program owners, and academic/career advisors. Ultimately this feature will be used with learners and made learner-facing.

Delivery Mechanism

To be confirmed: This feature will be delivered in the Skillabi interface as a new standalone page/function.

Success Criteria & Metrics

Definition of Done (PI-3): Research.

  • Identified approach to pulling groups of courses with similar skills, aligned to occupations, for a learner to select among

  • Confident that the approach to calling up relevant courses based on skill includes those that are most relevant and appropriate

  • Mechanism for calling up a series of possible course/program options, in order, for a learner to consider

  • Mechanism for calling up a series of similarly-leveled course/program options (i.e., intro or advanced courses) based on similar skills that relate to a career, for institutions to use as a starting point for new program creation

Dependency: 

Success metrics, measurable upon implementation (PI-4 and beyond):

  • TBD - need to determine implementation of feature in order to know how to measure.

Aspects that are out of scope (of this phase)

This epic focuses on research, and does not involve building skills-based learning pathways or implementing in the Skillabi tool.

 

PART 2

Solution Description

To investigate:

  • How might we cluster courses based on skill similarity, to provide learners with a list of possible courses to take to gain a skill?

    • Can we consider level/proficiency/sequence of the course when making recommendations to learners?

    • Similarly, how might we cluster courses to provide institutions with a set of related courses that could be recommended to form a microcredential/certificate/minor?

  • Identify what learning connections don’t make sense

    • Time constraints for students

    • % completed in degree influencing how big of a program change you might recommend – eg not recommending dramatic shift if you’re in your last semester 

    • Eg how do we set boundaries or requirements, similarly to career pathways – eg not recommending nurse if you’re a doctor as you’d need different specific/high level skills

  • What is the relevant intersection with career pathways; can we surface adjacent/similar careers based on curricular skill overlap?

 

Early UX (wireframes or mockups)

<FigmaLink>

Non-Functional Attributes & Usage Projections

Dependencies

Successful research would require a sufficient volume of training data using course content.

To implement a solution at scale, the following would accelerate and strengthen success:

  • Successful rollout for an institution would likely require institutions to provide the full catalog with confidence in the accuracy of the syllabi/course content materials uploaded & associated skill tags.

  • Successful rollout for an institution would also likely require connectivity to individual student records via LMS or SIS.

Legal and Ethical Considerations

No legal or ethical concerns.

High-Level Rollout Strategies

  • TBD

Risks

  • A learner would need to consider the course sequencing and prerequisite requirements when selecting courses to take. As a result, this solution would rely on either integrating with a course management system, and/or using as a supplemental tool with an academic advisor guiding learners (rather than the learner using independently).

Open Questions

  • How can we ensure that…

 


Complete with Engineering Teams

 

Effort Size Estimate

Estimated Costs

Direct Financial Costs

Are there direct costs that this feature entails? Dataset acquisition, server purchasing, software licenses, etc.?

 

Team Effort

Each team involved should give a general t-shirt size estimate of their work involved. As the epic proceeds, they can add a link to the Jira epic/issue associated with their portion of this work.

Team

Effort Estimate (T-shirt sizes)

Jira Link

Team

Effort Estimate (T-shirt sizes)

Jira Link

Models

M

 

 

 

 

 

Related content