Team Name:
Databaes
Team Members:
Databaes
Career Gameplan uses open data and graph database technologies to provide job recommendations and analysis on skill matches and gaps.
Our solution proposes using graph database technologies to identify the jobs that are most similar to the job they have now or have done in the past. This solution builds on the type of recommendation provided by joboutlook.com.au
to leverage the National Skills Commission’s Australian Skills Classification and create multiple different path recommendations, based on the skills that are known to be related and assumed transferable between skill clusters and families. This is then enhanced with data from applicable job listing and higher education course datasets, the descriptions of which are compared to skills text in the Skills Classification using machine learning methods, to determine the available jobs or education courses that are most semantically similar.
Description of Use Depending on availability, we would scrape or directly source job listing datasets like this one from the Moreton Bay Jobs website, to collect descriptions of jobs and job requirements, in order to do a similarity comparison to subsets of skills in the Australian Skills Classification.
Description of Use Based on availability of the source data, we would scrape or directly source a dataset containing course descriptions linked to courses on the course seeker website. The sourced data would be used for text similarity matching to the Australian Skills Classification, using the Skills-2-Vec library.
Description of Use This would be used to train a model that is able to compare and identify similarities between bodies of text in job adds or course descriptions, and a unique body of text created by the summaries of desired skills in the Australian Skills Classification
Description of Use We uploaded core attributes and mappings within the Skills Classification to a Neo4j graph database in order to calculate most popular destination jobs based on presence of linked paths between tasks, clusters and families.
Go to Challenge | 25 teams have entered this challenge.
Go to Challenge | 25 teams have entered this challenge.
Go to Challenge | 9 teams have entered this challenge.
