Trust metrics wiki - Trustlet:Community Portal


 * Here you can leave questions, comments, remarks for active contributors of this wiki.

License: to sharealike or not to sharealike, and template for papers

 * moved to project:License

The logo
There's not even a mention of Trust Metrics Wiki in the logo right now. I can ask amylin if she can produce something nice. Do you have some other ideas, or sketches or anything you can upload here? Guaka 13:05, 17 June 2007 (PDT)
 * I guess the logo will be different depending on the domain we decide to get. At the moment I totally have no ideas. I think it should have some network in some form. Inspiration can come from http://www.visualcomplexity.com/vc/ or http://infosthetics.com/

Trustlet Wikipedia and Others
If we want to compare trust system, we have to organize a base concepts of trust. I found authors calling in different ways the same thing. Many terms are present in wikipedia, others are not or they have a different meaning in the context of trust metrics. Should we link wikipedia voices? --RomualdoGrillo 10:07, 20 July 2007 (PDT)


 * If Wikipedia articles have evolved a bit they're probably using the concepts and terms we want to use here. Check the history of the articles, if there's been a lot of work on it you can probably trust it ;) guaka wikitalk 09:08, 21 July 2007 (PDT)

I added an article about Wikipedia vs Trustlet. I tryed to better define the limits of trustlet goals, in comparison with other open projects (wikipedia). I think when an overlap between open projects exist, the projects should merge, add each other, link each other. Fill free to modify and eventually cancel this voice (as usual).


 * I changed a bit the definition of trustlet, check if you agree. And thanks for the clarifying page (as usual) ;-) --PaoloMassa 04:40, 26 July 2007 (PDT)

I will add trustlet in the wiki list in wikipedia http://en.wikipedia.org/wiki/List_of_wikis --RomualdoGrillo 09:18, 26 July 2007 (PDT)


 * great idea! --82.61.133.128 11:17, 26 July 2007 (PDT)

Datasets, Databases, Metadata
When you begin talking about collecting and hosting datasets for research integration and analysis, you wind up with a number of categories of serious questions:

Technical Questions:

Host

 * what database program(s)/servers/scriping will you be using
 * MySQL?
 * Apache?
 * LAMP
 * spreadsheets

''At the moment, we are using a shared server on dreamhost. It has mysql, apache, gnu/linux. But people who will download software and datasets will be able to use what they prefer. The datasets will be exportable as simple text and possibly spreadsheet (opendocument format).'' --PaoloMassa 08:42, 17 August 2007 (PDT)


 * does the host support them and to what extent
 * how much room will each database take up and what are the size limits (virtual,physical,financial) at the host?
 * can the host scale up storage limits and user bandwidth to the anticipated needs of TrustLet

''There are no limits at the moment. We have a lot of space, 400 gigas or something like that and datasets will be text mainly so they will never occupy too much space.'' --PaoloMassa 08:42, 17 August 2007 (PDT)


 * Database management
 * who will be doing it
 * conversion of datasets to the database
 * conversion of the datasets to 'trustlet standard dataset elements'

''We need to decide the Network file format. Anyway I think it will be something very simple and text based. Text files will reside somewhere on the filesystem of trustlet.org'' --PaoloMassa 08:57, 17 August 2007 (PDT)

''Conversion between different formats is something we have to do if we want to serve the same datasets in many formats. Anyone willing to take care of this?'' --PaoloMassa 08:57, 17 August 2007 (PDT)

Purpose of inclusion
 * Selection of datasets
 * which datasets are appropriate for the primary goals of TrustLet

''The list is kept at Trust network datasets, we can be very broad in what can be there. At the very worst they will not be used... ;-)'' --PaoloMassa 08:57, 17 August 2007 (PDT)


 * which datasets can TrustLet responsibly manage (can they, or their users, trust US...)
 * especially in an open environment
 * do they include personal information that could be used for ID theft, trust system attacks, blackmail, etc.

''In general datasets that are public don't include this kind of information, so we don't have to worry about this I think. But it is important to think about it. Possibly we could anonymized all the datasets that contain personal information.'' --PaoloMassa 08:57, 17 August 2007 (PDT)


 * which datasets include sufficient metadata for accurate analysis/manipulation for trust metrics

the only information that is needed is some relationship between nodes, the more data the better but it is really hard to get anything too complex. --PaoloMassa 08:57, 17 August 2007 (PDT)


 * which datasets are reliable in their collection of data and metadata (which can WE trust...)
 * which datasets are secure enough to maintain a reliable dataset(can we trust them not to be hacked)
 * which datasets -can- be converted to the TrustLet initial or standard database and dataset elements

Most of the discussion about this is on the page Trust network datasets, there are not too many answers but a lot of questions --PaoloMassa 08:57, 17 August 2007 (PDT)


 * guaka wikitalk moved something to Dataset metadata

Standard TrustLet Data Elements (ok, I'll coin an acronym here: STLDE)
 * typical stuff here relative to definition of any standard data elements - type, length, range...

Security of the datasets hosted - (how well will we execute the trust of the data providers and the researchers)
 * host security
 * database security - database program modification,scripts,inputs,outputs,availability
 * dataset add/change/delete authority
 * dataset read permissions

At the moment there are not too much problems, data will be under a permissive license so other people will make other copies, and we will make backups on different machines. --PaoloMassa 09:21, 17 August 2007 (PDT)

I'm sure I'll come up with more later... d:D

Probably need a Glossary also. Tinker 17:38, 21 July 2007 (PDT)


 * Yes we need a glossary ;-) --PaoloMassa 09:21, 17 August 2007 (PDT)


 * Wow, these points are great! Thanks! At the moment I don't have time but I just wanted to say how much I appreciate your thoughts here (and actually all over the wiki!). We already thought a bit about this, for example about the Datasets license for the datasets and the "copyrights" of the data. But in the coming says we will try to discuss these questions and collaboratively come up with answers. Thanks once more! And yes we need a Glossary ;-) --PaoloMassa 23:54, 21 July 2007 (PDT)

Moving stuff from old wiki
We could also move some content from http://moloko.itc.it/trustmetricswiki/moin.cgi but not automatically, only what makes sense and reorganizing it in categories, templates, etc ...


 * We should promise ourselves to move at least one page a week, just to clean up there :) guaka wikitalk 13:58, 29 July 2007 (PDT)

New layout, and new sysop
I did some more work on the layout, courtesy of Lemon-head (who did something similar for Crash). I hope you like it. See also MediaWiki:Monobook.css and MediaWiki:Sidebar.

I also upgraded Callum to sysop. I like the work he has been doing here, he has a bit of experience with MediaWiki, and a lot with CSS, and I found out that trusting people on wikis usually leads to more activity :) guaka wikitalk 12:09, 24 September 2007 (PDT)


 * trusting people on wiki is an entire branch of research which would be great to study by itself! ;-) --PaoloMassa 06:40, 25 September 2007 (PDT)