Deal all,
 We would like to announce the release of a new update version of our
services with lots of bug fixes and new functionalities. Among other new
features, we provide a new training algorithm: Naive Bayes
Classification. We also provide a query through which a client can ask
for a certain 'Page' of given size from a uri-list. Here is an example:

http://opentox.ntua.gr:3000/model?page=2&pagesize=10 

this will return the second page of the results, given that the size of
every page is 10. We didn't have much time to write detailed
documentation for the services but we plan to do that by the end of next
week.

Here are some examples with curl:

1. Train a Support Vector Machine model for classification:
 curl -X POST -d
"dataset_uri=http://ambit.uni-plovdiv.bg:8080/ambit2/dataset/9" -d
"prediction_feature=http://ambit.uni-plovdiv.bg:8080/ambit2/feature/12136" http://opentox.ntua.gr:3000/algorithm/svc 

2. Train a naive bayes classification model:
curl -X POST -d
"dataset_uri=http://ambit.uni-plovdiv.bg:8080/ambit2/dataset/9" -d
"prediction_feature=http://ambit.uni-plovdiv.bg:8080/ambit2/feature/12136" http://opentox.ntua.gr:3000/algorithm/naiveBayes

3. Get the representation of an algorithm in PDF format:
curl -H 'Accept:application/pdf'
http://opentox.ntua.gr:3000/algorithm/naiveBayes > /path/to/file.pdf

We have also implemented a data cleanup service which removes features
of certain type from a dataset and returns a "clean" dataset uri the way
we had discussed that with Nina, but we'll deploy that by tomorrow.

Source code at http://github.com/alphaville/yaqp-turbo .

Best Regards,
Sopasakis Pantelis
Charalampos Chomenides