Cardbox Talk


CardboxForumsCardbox Talk > "Help with searching Cardbox databases"

Help with searching Cardbox databases

We store large volumes of data (CV's) and need to find a more efficient way of searching

Posted By Post


11-Jun-2008 10:21

We are a recruitment agency and receive a high volume of cv's, which we store in a Cardbox database - one cv per record, separating out the name. We realise we could do more to separate out different sections of the data (such as most recent job etc) but this can be time consuming due to the large amount we get in every day. Has anybody created anything within Cardbox that could be used to get to data in a more efficient manner ? Does anybody use a 3rd party tool to find data within Cardbox ?


12-Jun-2008 08:39


Your mail is amazing. An easier way to query a database I cannot remember...

I think you need to structure your input. Garbage in - garbage out...!


Charles Welling

12-Jun-2008 11:50

Bert is absolutely right. You cannot expect software to do the work of a trained information specialist. You could try to design a format in which you can store an abstract taken from from the CV's. Such an abstract should at least contain the fields of expertise of the applicant and the number of years of experience. Information of that kind can be stored using drop-down lists or checkboxes for fast and consistent entry. As CV's tend to vary in depth and in length, as well as in degree of exaggeration, they are not very well suited as a source of information unless they are properly processed.

Please note that the time that you invest in processing CV's hardly ever outweighs the time lost in laboriously trying to find a CV, not to mention the time and money wasted if you do not find an applicant who is in your database after all.

You do not say how you store the CV's. Do you copy them by hand or do you use OCR? If you should decide to structure the data, you could add the CV's as a scanned image which would save you a lot of time.


12-Jun-2008 17:15

Hello Sue,
As Bert and Charles have stated - the efficiency (speed but also relevance) of retrieval depends primarily of the way your data is structured.
The degree to which you can structure your records depends of the way the records are entered in the database - and that depends a lot of the format in which the CVs are submitted - in print or electronically.
Either way, you should, as Charles suggests, have, at a minimum, fields that are relevant to the way you identify candidates for a particular job, profession/positions held, level of education, years of experience (you could use the sample database "Applics" supplied with Cardbox as a starting point/inspiration). Because they are usually expressed in standard terms, It is easy to set up lists, check boxes or option buttons to make data entry fast and easy.
What is, I think, the most important search criterion in your situation, is the total of skills (technical and general) as they usually appear in the "job descriptions" or "job requirements". This may also be the most time-consuming to enter, but it can be made relatively easy, if you are willing to invest the time to create a "controlled vocabulary" for consistency, i.e. generate a list of standard terms /grammatical forms representing all the skills that may be required for any of the positions in your "inventory", and use it as a "pick list" during data entry (the list may be easily edited as terminologies or the professional profiles of your client base change)
Your inquiry does not state clearly the format in which CVs are submitted, although, from your statement "one cv per record, separating out the name. We realise we could do more to separate out different sections of the data ..." seems to indicate that you may be getting the "data" electronically.
If the CVs are submitted electronically (in North America this format has, I think, become prevalent), you can automate the skills indexing, by creating a field to import the full text in and apply the "index terms" validator linked to a "skills" list. In this case, the skills list, must be "richer" than the one I talked about earlier, to include synonyms and grammatical format variations for terms, as they may be used by different candidates in their CVs. This procedure is less precise, but much more time-efficient.
Of course, the format of the CV will be lost, but it is irrelevant for indexing purposes, and you can have another field for the CV, and use OLE to link to the original document, or a picture field to insert pages in PDF format.
I hope this helps. If you have any further questions, please feel free to e-mail me directly to icar-consulting@rogers.com I will be happy to help (my reply may be delayed because of the time zones).


9-Oct-2008 10:01

I have a similar situation. We'd like to create a database of our available digital information. We'd like a search function which isn't limited to looking at the contents of record fields, but actually searches INSIDE the attached document itself. Does Cardbox offer that possibility and if not, are there any other sources you can suggest?


9-Oct-2008 10:25

First you have to copy the content of a doc into a Cardbox Field. When you query Cardbox you'll find all records/documents containing you search term.
To find the search term in the relevant document, you have "to macro" something. You can start that macro by a screenbutton and the search can be done in Word or Acrobat doc etc.
What's neeeded?
- The (full) documentname in a field of the record
- A macro which finds the entered search term(s) in Cardbox (from the search history perhaps)

(I one application I also did the whole selectionproces by a macro. In that macro also the search terms were stored into ScrapFields. The content of the Scrapfields could be used for searching on same terms in Google by pushing one button).

Bert S.

© 2010 Cardbox Software Limited   Home