[Sigia-l] Entering CVs online

[Gwen Zanin]

We have a resume bank on our education recruitment site where users can 
enter and manage their CVs for online applications.  We intend to make 
these resumes searchable for our advertisers in the near future.

We also wrestled for a long while over the barrier-to-entry question and 
the trade-offs of having discrete fields (making for nicely searchable 
and reformattable data) vs one big glob of text (easier to enter but 
harder to search).  We decided we didn't want to turn people off from 
having to fill in so many questions and so many fields -- partly due to 
business reasons and wanting to have a nice large set.  This trade-off 
will certainly make search results a lot less precise, however.

I tried out a few of the products you mentioned in your email but I was 
very unhappy with a lot of the incorrect assumptions I saw in the 
results as far as tagging the data went.  Have you or anyone else found 
an automatic parser for CVs that you're happy with?

Gwen Zanin
Senior Database Designer
The Chronicle of Higher Education
> ------------------------------
> 
> Message: 4
> Date: Thu, 03 Feb 2005 14:17:22 -0500
> From: David Crow <sigia-l at davidcrow.ca>
> Subject: Re: [Sigia-l] Entering CVs online
> To: sigia-l at asis.org
> Message-ID: <420278C2.7090101 at davidcrow.ca>
> Content-Type: text/plain; charset=ISO-8859-1; format=flowed
> 
> The presentation Weston mentions is available online from EDUCASE at: 
> http://www.educause.edu/ir/library/powerpoint/EDU0326.pps
> 
> We've been moving towards HR-XML resume format for storage (structured data, it has some advantages when we move from the custom system to a Tier 1 ERP). It is an incredibly flexible method for storage. 
> 
> But I keep fighting should I make all of my users enter their resume data in these descrete chunks. It is a lot of work for users to change their resumes into a chunks that fit my storage model. Our informal user feedback and usability tests show that our users don't want to enter fields and fields of data (this might be related to our user group). The overhead required to enter your resume in a structured format might eliminate a set of potential hires (applicants with cognitive disabilities; applicants who really haven't decided to leave their current jobs). 
> 
> Currently, we allow users to submit their CVs as text, rich text, word, pdf, etc. This is stored as a clob in the database. Non-structured data. It's horrible (but they print nicely).
> 
> I've been looking at using products to parse the resume into HR-XML format including: 
>  * http://www.resumeparser.com/
>  * http://www.resumemirror.com/products/resume-processing.html
>  * http://www.greenhire.com/GHParser.html
> 
> Each of these products would allow our users to upload their resume in a file, and we can then create the nice HR-XML structure without the user having to know. We then add the file to the resume using the <SupportingMaterials> tags. It significantly reduces the effort and improves the customer experience.
> 
> <Resume>
>   <NonXMLResume>
>    <SupportingMaterials>
>      <AttachmentReference mimeType="application/pdf">submitted_document.pdf</AttachmentReference>
>    </SupportingMaterials>
>   </NonXMLResume>
> </Resume>
> 
> These products are expensive, but I am more inclusive in the user base that I would be able to serve. And I would get the benefits and flexibility of a structure data format.