Contextual Resume Parsing Software (Follow-up Query)
I'm still searching for anyone who can shed light on the "next generation of resume management software."
An article written many, many months back [don't remember by whom, and unfortunately, can't find] outlined how resume parsing software uses a points system to analyze and grade resumes. A jobseeker holding a bachelor's degree or a certain amount of work experience for example would be given points. Those points were then calculated and the software would present the best candidates in the system based in part on a mathematical calculation.
Can someone elaborate on what companies are using these days? What new features do resume management systems now offer hiring managers?
Regards,
@teenarose
An article written many, many months back [don't remember by whom, and unfortunately, can't find] outlined how resume parsing software uses a points system to analyze and grade resumes. A jobseeker holding a bachelor's degree or a certain amount of work experience for example would be given points. Those points were then calculated and the software would present the best candidates in the system based in part on a mathematical calculation.
Can someone elaborate on what companies are using these days? What new features do resume management systems now offer hiring managers?
Regards,
@teenarose
Tags: management, parsing, resume, software, system
Comment
-
Comment by Brian Busby on January 26, 2010 at 7:43pm - We are working on this very thing right now. Our system will "learn" on a daily basis what exactly is being asked of it. As part of this overhaul we are scraping our current search engine entirely and developing a system employing integrated technologies of automatic concept extraction/matching from text, a powerful fuzzy search engine, and a collaborative user preference learning engine to provide highly accurate and personalized search results, similar to the Trovix engine. In this method the system will maintain unique information for every search from each user to deliver results that are the most appealing and applicable to that particular user. The knowledge base will allow the job seekers resume to be stored and retrieved in a way that the relationships in the data will have relationships with all other relevant data, categories, industries, etc, not simply pulling keywords from the document to return. Text query simply cannot learn or return data with near the level accuracy because of misspelled words, abbreviations, missing spaces, etc. as well as no understanding of how the data relates.
-
Comment by Teena Rose on January 26, 2010 at 11:58am - Thanks for the response, Dennis.
-
Comment by Dennis Gorelik on January 26, 2010 at 11:27am -
I think current mainstream practice is just to rely on keywords match.
It could be done better though. For example, every keyword could have it's own weight (according to recruiter's preferences), so that way every resume could be rated based on recruiter's preferences.
Recruiter doesn't have to manually set weights on every keyword. They system can learn from day-to-day recruiters choices: if recruiter rejected resume, then every keyword in that resume would be downgraded a little. If recruiter liked resume then every keyword in that resume would be upgraded a little. After 20+ resume evaluation the system would have pretty good idea about what recruiter is really looking for and would present such resumes to the recruiter among the top results.
The system can keep learning.
I plan to implement system like that on postjobfree.com, but didn't have time to dive into that yet.
© 2010 Created by Slouch on Ning. Create a Ning Network!
You need to be a member of Job Boarders to add comments!
Join Job Boarders