Home Page

• Technology
• Data analysis
• Direct data editing
• facts sheet (pdf)

Demos

• Walkthrough
• Data analysis
• Direct data editing

Try ILOG Discovery

Use it in your application

 FAQ >

Contact

Frequently Asked Questions on:

• ILOG Discovery product
• Technical issues

ILOG Discovery product

How does a tool such as ILOG Discovery fits with business analytics, data mining and knowledge discovery solutions ?

Visual data analysis tools such as ILOG Discovery can be used either as a complement or an easy to use replacement for data mining solutions:

• As a complement, ILOG Discovery is useful in the early stages of knowledge discovery, to get a good picture of your data sets, cleanup data and find out where to dig for. During the process, ILOG Discovery is useful to view intermediate results, filter out findings that do not seem to make sense. The most prominent feature of visual data analysis is that it lets you introduce common knowledge into the discovery process: from your knowledge of the domain, you can often find explanations for obscure findings, relate the correlations to some knowledge which is not in the data, but that you know well. Hence it is useful to drive the knowledge discovery algorithms. Finally, ILOG Discovery becomes a powerful tool to report and share subtle findings: those need to be put into context and highlighted properly to be convincing. Because ILOG Discovery lets you tailor very finely your presentations, you can choose the visual representations that will best highlight what you are trying to say.
• As a replacement, visual data analysis is as easy to use as pivot tables. It does not require data cleaning, provides many useful ready-to-use visualizations, which often let you immediately spot patterns, exceptions or get a sense of how data is structured. While not providing you with statistically validated results, visual data analysis provides often telling results without requiring mastering sophisticated data mining tools.

I have the preview version. I don't understand such or such feature, OR there seems to be a bug when I do this or that OR I wish we could do this...

We provide limited user support and welcome all feature requests. Just contact us.

In your description, you say that ILOG Discovery is a component. What is its API and how can I evaluate it ?

ILOG Discovery is both an interactive tool and a component. In the current preview, we propose you to test only the interactive part of ILOG Discovery, even though the component is available to selected users for evaluation purposes. For instance, the ILOG Visual CP extension is one such use we foresee for the component. If you want to use ILOG Discovery features in your application, see the API page.

How much will ILOG Discovery cost ?

Pricing is not determined yet. Most likely, the interactive tool will come at a set price, while a development version enabling connecting ILOG Discovery to enterprise information systems and deploy views on enterprise portals will come at a higher price. 

How did ILOG Discovery start ?

ILOG Discovery started as a research project to provide debugging tools and visualization of constraints-based programs. As such, it is partly funded by the French ministry of research RNTL, within the framework of the OADymPPaC research project.

I would like to share my experience, or get to know examples of use.

First, start by registering your email address with us. We won't spam you, but will let you know of newer versions, samples and documentation as they come out, sending you at the very most 4 emails. We will soon provide a forum on this website, where you will be able to exchange examples of uses, tricks hints and applications of visual data analysis for various domains.

Technical issues

I get the message "License file contains invalid keys", or "License key expired" when starting ILOG Discovery.

We are sorry about the inconvenience these keys sometimes generate, however, we need to keep some kind of track of how is ILOG Discovery used to provide the best possible product. These messages can appear if you have entered a wrong key, or used a wrong machine name to generate the key, or changed your machine name by changing the network settings. To fix all key related issues, no matter what the issue is: 

1. Erase or rename the license key file. This file is located at: (Windows: Documents and Settings\[Your login name]\ILOG\java\discovery\Keys ; Unices: $HOME\.ILOG\java\discovery\Keys).
2. Start ILOG Discovery. The program will indicate it requires a key. Press OK without entering a key. A new dialog will appear that will indicate the actual machine name the program is expecting for generating a key. Note this name carefully.
3. Go to the registration page (http://www2.ilog.com/preview/Discovery/registration) and enter the appropriate machine name. You will receive a new key by email.
4. Start ILOG Discovery again, copy/paste the key you received, and only the key: it is a string of 35 characters and digits, starting with the letter D.

ILOG Discovery should now run correctly.

I have installed a new version and the new version won't run, or when I double-click on a project file, the old version gets launched.

This can happen, specially on the Mac OS version. On MacOS, you need to remove the previously installed version, to ensure double-clicking on a project file always open the latest available version. Besides this issue, preference files may not always be fully compatible between versions. It is always safe to remove the old preference files after having installed a newer version. This file is located:

• on Unix and MacOS, in the user's home directory, under ~/.ILOG/java/discovery/discovery.prf
• on Windows, in the user's home directory, under Application Data\ILOG\java\discovery\discovery.prf

I don't find the parallel histograms any more ?

As of version 1.07, we had to remove the parallel histograms to avoid potential legal issues. The other types of overviews provide as much, if not more, insight on your data sets.

Known issues, bugs, limitations and feature requests as of 27/11/04

This list presents roughly the bug reports and feature requests we have received so far. Those are processed on a case by case basis.