Applicant Tracking Candidate Matching Advancements
Automated candidate matching within applicant tracking systems (ATS) remains one of the most attractive features for recruiters that deal with a high volume of candidates for job requisitions. The technology behind the matching continues strengthen and new algorithms deliver higher accuracy. However, just like every other technology, it is not 100%.
In my last position, I had the opportunity to view VURV’s matching capabilities in a demo. However, it really boiled down to a search function that provided you with a generic ranking of candidates. For the demo, the presenter searched by the keyword “sales” in an effort to find a new candidate for a sales position. The results listed the candidates in a ranked listing with instant access to their resumes. When it came time to review the resume, it was obvious the resume had been created in a SEO format because the highlighted keyword seemed to be every other word. Congrats to the candidate, but was he right for the job?
Today, there are companies that focus strictly on providing vendors with candidate matching functions that plug directly into an ATS. One of the most promising candidate matching engines is being developed by Sovren Group. They offer the Sovren Semantic Matching Engine, and the technology behind the matching is truly amazing.
In the future, I would be willing to bet that ATS vendors will start offering a percentage guarantee of accuracy for candidate matching. Many years down the road…of course.
Filed under: Applicant Tracking
Tags: Applicant Tracking Candidate Matching, ATS Candidate Matching, Resume Matching
Hi Jake,
Nice to read your opinion on what matching can do.
May I invite you to take a look at our website (www.actonomy.com), which is strictly focussing on HR-data matching which we integrate into theird parties’ ATS, e-recruitment systems, and job boards.
As you said, there are still ATS systems which claim they can “match” data, but which are in fact only searching for keywords, which is a big difference!
Good (but unfortunately not all) matching technology offers: semantic matching, fuzzy matching, reversed or bi-directional matching, asymmetric matching, weighted matching, CV to CV matching, jobs to resume matching, guided matching, auto suggestions, ranked results with detailed scoring results explaining why a resume matches, etc,
It’s remarkable to see that many ATS systems still have underperforming search & match functionality. They do not use modern matching techniques helping recruiters to work better and faster, but they still use database query techniques, or simple search appliances such as google enterprise search, which lack the speed and the artificial intelligence recruiters could use to perform optimal in their daily search for candidates, or candidates in their search for jobs. It’s often a reason for recruiters to not search their database sufficiently, and spend money advertising jobs externally!
We have just launched an extra tool/feature which is integrated into our matching engine: The largest HR knowledge base for recruiters, by recruiters: it knows relations between words in CVs, and words recruiters use to search. It contains expert recruiters knowledge, which is automatically applied when you search through databases (in ATS’s, or job boards, etc).
Example: let’s say you’re looking for a “java programmer”. What if an unknown candidates’ CV mentioned “Object Oriented Programmer”..?
In that case, in traditional systems, it would not appear in the results. The HR knowledge base solves these problems: it contains an ontology which knows the relations between words, and in this example it knows that Java is an instance of oriented programming, so it WILL find that candidate’s resume, and show it in the results list, but with another weight than a 1on1 matching “java programmer”. The same goes for e.g. “Key Account Manager”: it wil also automatically search for people with “Sales Manager” in their CV, but will assign another weight and ranking in the results list. It’s even possible to search for a function, based on competencies and skills. This also applies to candidates who search for jobs. The problem is also here: “what if the recruiter used different wording than those the candidate uses to search for a job…?”.. The HR knowledge base automatically supports searches just like expert searchers would do.
Pls don’t let this sound as a sales pitch. It’s just to inform you about what matching CAN do..
Let me know if you need extra information. Pls feel free to send me your email address (mail to info@actonomy.com, and mention this blogpost) so I can send you a recent press release on this.
Kind regards,
Bart
http://www.actonomy.com