Guess your next job.

By applying data mining techniques to reed.co.uk's databases, we can try to predict the next job in a career path and help people to better understand their career options. Data mining techniques such as Information Retrieval (IR), clustering, and association analysis are used together to uncover interesting hidden data patterns. GUESS YOUR NEXT JOB an amazing tool from reed.co.uk by Dr. Fernando 1. THE PROCESS 7 (from data-mess to clear conclusions) 325k candidate CVs DIRTY DATA CV data (job history) Posted jobs Candidate data CLEANING 250k 1.5 m 260k 650k unique job titles CV entries unique job titles posted jobs Information Retrieval UR) Processing REDUCTION Remove inconsistent/ missing data IR processing Data intergration/ selection Data selection 150k job titles GROUPING Clustering Text Mining 1.2m CV records 150k 500k 40k 300k CLEAN DATA job titles job titles jobs candidates Data needs to be pre-processed as dirty data can cause confusion for the mining procedure and increase the time and complexity of the process. To process documents, IR is used while clustering is used to group similar job titles. 15k Salarkes by title Stor Preferred 9ators/ job titles Dkmographie Association analysis is used to generáte rules which reflect job titles that are torether andtn represent frequent patterns in the job sequences. Sequence Analysis applications • Recommendation systems for both recruiters and candidates • Search engines Personalization • Demographic indicators CAREER PATH MACHINE ! 2. CAREER PATHS / detis see some practical cases) €) = average salary E- average time in that position = % following that career path = average age ACCOUNTANT Financial Controller Finance Director 16% 40% 35k W 29 months 50k F 32 months € 73k 41 months 31 years old 36 years old O 39 years old 23% Head Waiter Restaurant Manager 19% WAITER 4% 18k H 26 months € 24k 9 30 months General Manager 24 ysırs old 27 years old 5% Bar Manager 3% € 15k 16 months Bartender 8% £ 32k H 33 months 20 yaars old 33 years old 16k 19 14 months 22k a 26 months 21 years old 24 yeırs old Iam currently a software developer looking to progress in a path in which the development of my technical abilities is more important. 23% Technical Lead Senior Developer € 52k H 21 months 30 yeırs old 25%) DEVELOPER € 44k 23 months 19% 30 years old Consultant € 50k H 19 months 15% 36k E 18 months 33 years old 29 years old 4% Business Analyst Project Manager 28% € 46k H 23 months € 40k H 18 months Careers can often follow different paths; I would like to specialize in a managementlanalysis role. 31 ysırs old 34 years old B. DERIVED DATA / other interesting stuff) INFO +KNOWLEDGE 10% 4 28 23 average numbeR of jobs candidates have had of candidates have had six OR ONLY MORE jobs average time candidates have 25 average time candidates MONTHS change jobs MONTHS the same job title age with the highest number of job applications 20% jobs in CVs have lasted more than ONLY 3 years 21 average age < candidates have their fIRst job 3 MORE popular titles in CVs ORE Sales Assistant, Administrator and Receptionist O People change jobs more often when they work in sectors like: Apprenticeships (8 months), Graduate training (13 months), Marketing & PR (21 months) People stay in a job for longer when they work in sectors like: Manufacturing (38 months), Engineering (35 months). Data mining provides tools and techniques that can extract previously unknown interesting patterns and find dependencies in data. This new knowledge can be used not only to develop helpful tools for both job seekers and recruiters but it can also provide reed.co.uk with a better understanding and insight into the job market. reed couk The UK's #1 job site.