we have had many discussions in the past hosting on TransScript Area on Chatbots.org. Now, inspired by the thread by Wouter, http://www.chatbots.org/ai_zone/viewthread/1459/, I’m happy to inform you I do want to have such an area on Chatbots.org, provided this area will be build be the community:

Here are the specifications
[ul]
[li]Sub tab under directory: TransScripts (main tab will be renamed into ‘virtual humans’ with a subtab ‘gallery’)[/li]
[li]Left column should contain a list to all transscripts, a way to navigate through all the transscripts (rating???) [/li]
[li]Button: Upload your Transscripts (20 contribution points)[/li]
[/ul]

On uploading, the following meta data need to be entered:[ul]
[li]Chatbot name (or be selected from a dropdown if the chatbot is listed on Chatbots.org)[/li]
[li]Recording date[/li]
[li]Contest (dropdown, optional)[/li]
[li]Year (if contest is selected)[/li]
[li]upload file. Which formats should be allowed? At least .txt? Anything else?[/li][/ul]

Uploaded transscripts should be approved by one of our moderators. Dave, Steve, me or anybody else who want to volunteer in this area. We should make use of our standard workflow mechanism to upload and moderate content (like the ‘add your chatbot’ button).

Once uploaded, the transscripts should be displayed. We should interpret the transscripts intelligently and format the lines by the conversational parties involved in a different way. (different backgrond color for example).

In the directory tab, i.e. the chatbot listings, and in the awards tab, the transscripts will be included automatically.

This should be the basic functionality. The forms and workflow are pretty Chatbots.org standard, and our VPS has more than enough space, but the formatting is a bit more difficult.

Wish: a rating mechanism for transscripts. Shouldn’t be that difficult…

In a next phase I’d love to introduce ‘commenting on line level’, so people can start a discussion on line 241 in the transscript in a very easy way.

Which php programmer would love to work on this?

Looks solid as a plan.

One early technical decision that should be made, is how to store the different transcripts. Plaintext files? In a database?

I would argue for a separate plaintext file for each individual transcript, and perhaps stored in folders by year. This might quickly run into the thousands, but I think that would be an acceptible tradeoff for the many advantages: easy parsing and creating, a ‘timeless’ format that can be handled by any programming language (and even, say, Excel), and a solid conceptual ‘individual’ log you can point to, easy creation of possible ‘bulk download’ options.

Mind you, I am not saying they should be *displayed* as text files. A link to a transcript should take the relevant file, and then show that text in a nicely formatted way, like you say, with background colors and things like that.

I have the impression that this site is driven by the Joomla! platform, is that true?

I am a php programmer and wouldn’t mind giving this an early shot - is there a way to create a ‘hidden’ page and then grant me access to it (and an FTP account that can upload php files and transcript files to test etc…)

Another question: would there be any ‘copyright’ issues with hosting transcript files rather than just linking to them online? For example, would anybody object to chatbots.org hosting Loebner transcripts? (with proper attribution of course)

nope, it is created with http://www.expressionengine.com

We would create CMS entries for the meta data, and have separate text files to upload….

Would you like to experiment a bit? I can give you access to our development server…

Sure, why not - feel free to send me an private message (you can do that through this forum, right?) with the details, credentials, and any more thoughts you may have on this besides this thread (though you laid out a pretty comprehensive plan already I would say), and I’ll see what I can do. As a chatbot-enthusiast it should go without saying that I love experimenting

The only way to send “private messages” here is through the site’s email system, using the links below our respective avatar images. I suppose that’s the same thing, semantically, but since that “leaves the confines of the forum”, it doesn’t really amount to the same thing, to my way of thinking.

Ah yes, that’s what I meant - I’ve gotten email like that before from fellow forum members - Erwin, feel free to use that channel!