- Her background – how she got to where she is
- How skills are handled in Coursera's skills graph
- Current applications, including how the skills graph was built and how skills relate to finding relevant content
- Where skills are going next
Some of the questions asked and answered in the webinar include (more questions answered in the webinar can be found here):
- Do you collect the tag info "is taught by" in what you will learn about this course? Some courses have this info under "About this Course."
- Do you have your own search engine or do you use something like solr? was it difficult to add your new module to existing infrastructure?
- Re: R vs Python: I teach Python, but have no experience of R. I read a recent article that suggested that R was basically dead, and that Python was the way to go. Is that overstatement?
- Do you think any of the work that you've done can translate to non-Coursera communities? Specifically ML developers?
- Do you have a knowledge engineer on your team? How do you know that there is no better ontology for your purposes?
- How important is understanding the business domain for data scientists? How much knowledge data scientist have related to business analysis to identify and work on different business problems?
- How many tags do you have? What does the classifier that predicts them look like? What metrics do you use to assess them?
- Have you ever looked at trying to analyse the learning style of learners ie preference to reading/videos/quizzes/assignments. Could identifying ratio of reading to videos to quizzes to assignments etc. help with recommendations?
About the Speaker
Rachel Reddick received a PhD in astrophysics from Stanford, but realized she wanted to do more down-to-earth work. After some exploration, she switched into data science. She first worked as a data scientist at Bosch, a large manufacturing company. Afterwards, she joined Coursera, where she's been working for almost two years.