References and Appendicies

References

  1. National Skills Commission (2020), The shape of Australia’s post COVID-19 workforce, Commonwealth of Australia, Canberra, Australia.
  2. Joyce, S. (2019), Strengthening Skills: Expert Review of Australia’s Vocational Education and Training System, Commonwealth of Australia, Canberra, Australia.
  3. World Economic Forum (2018) Towards a Reskilling Revolution: A Future of Jobs for All, World Economic Forum, Geneva, Switzerland.
  4. Department of Employment, Skills, Small and Family Business (2019), Reskilling Australia - a data driven approach, Commonwealth of Australia, Canberra, Australia.
  5. Kern, M.L., McCarthy, P.X, Chakrabarty, D., & Rizoiu, M. (2019) Social media-predicted personality traits and values can help match people to their ideal jobs, Proceedings of the National Academy of Sciences, Melbourne, Australia.
  6. Google (2020), Embeddings: Translating to a Lower-Dimensional Space, Google Developers Site: https://developers.google.com/machine-learning/crash-course/embeddings/translating-to-a-lower-dimensional-space
  7. Mikolov et al., (2013), Efficient Estimation of Word Representations in Vector Space, Cornell University, arXiv:1301.3781
  8. Brown et al., (2020), Language Models are Few-Shot Learners, Cornell University, arXiv:2005.14165  
  9. Devlin et al., (2019), BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding, Cornell University, arXiv:1810.04805
  10. van de Maaten, L. & Hinton, G., (2008), Visualizing Data using t-SNE, Journal of Machine Learning Research

Appendix A: Searchable Dashboard - Exploratory analysis of VET Qualification Similarity

Figure 4: Screenshot of the Exploratory analysis of VET Qualification Similarity dashboard displaying the overall similarity of Certificate III in Individual Support to other qualifications across training packages
Source: NSC website, 2021

Image
Screenshot of the Qualification similarity interactive appendix

The Exploratory analysis of VET Qualification Similarity dashboard is an interactive tool which compares the similarity of one qualification to all other qualifications in training packages in Australia.

The searchable dashboard is available on the NSC website.

Accredited qualifications (those outside training packages) are not included in this tool. 

The similarity of a qualification has been graded as Very High, High, Moderate, and  Low.

These gradings are based on cut-offs of the similarity scores to all other qualifications based on the overall similarity focus.

The top 20 matches for each qualification were segmented into percentiles for their similarity scores:

  • top 25 per cent of similarity scores graded as ‘very high’, 
  • second 25 per cent of similarity scores graded as ‘high’, 
  • third 25 per cent of similarity scores graded as ‘moderate’ 
  • remainder of similarity scores graded as ‘low’.

Filters can be applied to view the rankings of similar qualifications based on:

  • Title similarity
  • Description similarity
  • Keyword similarity
  • Core unit Similarity
  • Elective unit similarity.

Appendix B

Figure 5: The t-SNE algorithm produces a scatterplot showing a selection of training packages and how their qualifications relate to each-other in terms of similarity.

Image
The t-SNE algorithm produces a scatterplot showing a selection of training packages and how their qualifications relate to each-other in terms of similarity

The t-SNE scatterplot in the discussion section (figure 3) highlighted links in qualifications for the training packages community services, health, and defence. 

Figure 5 above presents a fuller picture of the qualification landscape by showing some of the training packages with the most qualifications.

As in figure 3, this shows how individual qualifications relate to each other in terms of similarity and how clusters of qualifications intersect within and across training package.