NEC AFIS problems?

[rsc55]

Hello all,

My agency currently uses two separate AFIS systems - One to search our county, and NEC to search California state. We typically search the local database first, and if no hit, we move on to the state system (as is logical). Over the past year or so, we have noticed prints hitting in state which have not been hitting using our local system, even though these subjects have been printed in our county. In an attempt to evaluate whether or not we actually need this local system if the state system is actually more reliable, my supervisor had me start running things in state first, and then in local if there is no hit. For about 2 months, I wasn't hitting on anything in local that was not hitting in state first. However, these past 3 weeks, I've had 4 prints not hit in state, but hit in local. In fact, 3 of them were in this week alone! This has happened on both palms and fingers.

My question to everyone is: has anyone else noticed a lower hit rate recently with NEC (maybe even in California)? It seems odd for state AFIS to all of a sudden become less reliable. I realize state is a larger database and so on, but for all of the "misses" to be clustered like this is suspicious.

Any input is greatly appreciated!

[Bill Schade]

This does not seem to be the appropriate forum for your question

NEC has a users group which should be a great source of information on a subject like this.

I would also suggest that you speak to your NEC service people to look forther into your inquiry.

Just a thought

[Ken Chappell]

You have the advantage of having a secondary system to compare your results by. Sortof a litmus test to gauge your NEC system to.
I haven't experienced what you have described and since our upgrade at the first of this year our hit stats did increase a couple of percentage points because of the added slap print search to our system. (25%-33% hit ratio) Our match scores have also increased too.

As Bill Schade suggested, the NEC Users group and NEC themselves might be able to answer your question much more indepth. Although I think your question is suitable for this forum because if you don't ask questions you will never get answers.

[johnab]

Good questions (and good advice Bill) ... and Ken

I am a latent examiner and do some training for NEC so feel free to call or write with any queries rsc55.

Will not be in the office Monday but will be available the rest of the week.
Regards

John Burt - email : john.burt@necam.com

Office phone (916) 4637025

[David Fairhurst]

My tuppenny opinion is that this may be down to differing quality in the exemplars between the two systems.
As for the illusion of clustering, it's just that (http://en.wikipedia.org/wiki/Clustering_illusion). Random events do not occur at regualr intervals, that just wouldn't be random.

[jasonmerlo]

If you are using a camera in conjunction with your NEC system make sure that it is calibrated correctly. If you are not searching 1:1 you are not going to hit anything (I'm assuming you're not doing 5:1 searches). You can test your system by searching known prints that you know are in the system and should hit.

I will say that this is just how it goes sometimes. My agency performs state searches via an NEC AFIS and then we move on to ULW for an IAFIS search if we don't hit in the state system. There are plenty of occasions when I hit people in ULW that had a record in our state, on the exact same knowns in the system. The ULW algorithm just catches stuff sometimes that the NEC algorithm doesn't.

[rsc55]

Thank you all for your helpful replies. I have since been in contact with johnab about the problem, and he has helped me greatly. It seems to have really been a random cluster of events. All I can say is, it's a good thing we have both systems! They seem to do a good job in conjunction, and I will definitely be recommending keeping both systems.