As a company finding the right candidate to fill a position is often difficult, time-consuming and hence costly. Also as a candidate the job search can be a very tiring process. There are already many software-tools that claim to make this process easier for both sides. However, one main problem is that there is no standard in writing job postings. Different companies use different words to describe the same job. These words might not be obvious to job seekers. That makes it hard to match candidates looking for a specific job by using search keywords with the actual open vacancy.
Google wants to solve this issue and offers its Cloud Job Discovery service to job boards, applicant tracking systems and related companies. The API offers access to Google’s search function and Machine Learning services and thereby claims to improve job discovery, job-seeker experience and conversion rates.
What Cloud Job Discovery can do
1. It filters out acronyms and synonyms so that when a candidate searches for a specific keyword it also searches for industry and company specific jargon.
2. It autocompletes job description with missing information like employment type, benefits, company address etc.
3. It allows job seekers to filter for commute times and preferred mode of transportation to better compare indirect benefits of different offerings with each other.
4. It recognises seniority levels and thereby displays only relevant openings to the job-seeker.
5. It allows the job seeker to indicate which job opening he likes and which not and thereby learns and tailors the search results to the individual.
To do so, Job Discovery takes all job entries from a job board or a career site and passes, classifies and maps the content to pre-trained data models.The data gets then codified into title, skills, location, education and many more. Behind each of these categories there is a data model (e.g. the “Title” data model contains 254.000 normalised titles) that ensures, with the help of Machine Learning, that the job seeker finds exactly the position he is looking for.
This is just another example of how Machine Learning can and already does improve the efficiency of a manual process. It surely is going to change the way of how people will search and find jobs even more in the future.
References
https://blog.goboon.co/posts/how-machine-learning-is-revolutionizing-recruiting
http://blog.gethappie.me/meet-googles-new-job-match-api-cloud-jobs
https://cloud.google.com/job-discovery/
https://youtube.com/watch?v=FvWyXQ9eg6
Hello Jannik,
The cloud job discovery sounds very interesting. It would be great to use machine learning for this. However, I believe it will have big impact on existing jobs, like recruiters. Also, this would be a threat for other online recruiting platforms, like linkedin. I wonder wether face to face interviews will remain to exist in the future. Some big companies are already using AI for their job interviews. I personally would prefer having face to face contact. I am curious how this will develop in the future.
Hi Jannik, nice choice of topic, I was just about to write about exactly the same! 🙂
I think what Google Jobs can do is well covered in your post. What would be interesting in a next step, would be to know how this is going to affect the recruiting industry. Google officially claims that they do not want to enter the recruitment market but by lowering the search costs they still affect it very much. If the algorithm performs as good as Google claims, it renders the traditional recruiting agency unnecessary as the job seekers and employer are directly connected. For the job board sites (say monster.com and the like) it could proof very valuable if they can be listed on Google Jobs and make their ads easily findable for job seekers. It would in fact even make their ads more valuable as there will be more suitable applicants seeing their listings. However, if Google lets employers directly post positions on Google Jobs this could mean the end of many job boards. It might be just a matter of time as this would be a very lucrative move…