Re: [Hypercup-devel] R: Towards a more deep investigation of JeromeDL !

[Sebastian K. <seb...@de...>]

-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

Dear Giuseppe,


On 24 Sep 2007, at 16:30, Giuseppe Bux wrote:

> Sebastian
> Here in the following my comments to the below question points.
>
> 1) Ok! I'm looking for.
>
> 2.1.1, 2.1.2) I tried to access bleedingedge.jeromedl.org but the =20
> service is temporarily unavailable. I'm looking for credentials to =20
> access it.
>

I have moved bleedingedge.jeromedl.org to a new, very fast server - =20
let me know if you will encounter any problems.
Probably the best way would be by some IM.
>
>
> 2.2.1, 2.2.2) I assume that Semantic queries are the main focus of =20
> JeromeDL: they should be allowed on the whole JeromeDL knowledge =20
> base, by using metadata from Marcont, FoafRealm and =20
> JeromeDL.structure ontologies. Templates for more frequent user =20
> queries, concerned with people and subjects of resources are =20
> necessary. It could be useful, however to provide an rdf query =20
> template enabling more advanced users to select metadata from the =20
> supported ontologies and give some value constraints (the Serql =20
> WHERE clause).
>
>

> 2.2.2)
> * Because of the semantic perspective of resource management, some =20
> NLP algorithms should be provided: at least in order to manage in =20
> the RDF queries, the semantic descriptions of domains and keywords =20
> available in the wordnet and taxonomies repositories.
> * As to the provision of more advanced NLP algorithms, such as the =20
> ones concerned with lexical analysis and word disambiguation in a =20
> given text context, I could talk about with researchers of the =20
> local Bari University. I will let you know about.
>
>

That was the idea of NLQ - but since that there was no time to put =20
any effort more into this component. Once I am done with current =20
implementations I will see what can be done to improve these =20
features. I will sure ask for some help from NLP people - since we =20
are inevitable going this direction with NLQ.



> 3) I'm looking for.
>
> 4) A more refined mapping is necessary to effectively communicate =20
> with other digital libraries.
>

Yes, the mms.marcont.org service is down at the moment - I am trying =20
to chase up the people responsible for this service. Maciej was =20
suppose to improve the MMS rules we use, but he did not have time =20
before his holiday - I hope this will be completed by the end of =20
October.

>
>
> 4+) I assume the Hypercup web service  could act as the jeromedl =20
> gateway to communicate both with other digital libraries and with =20
> external third part services platforms providing add on features =20
> (ex. multichannel/multi-device resource distribution, resource =20
> composition and printing on demand, e-commerce transactions, etc.). =20=

> My colleague Luigi Intonti is in charge of investigating hypercup: =20
> I add its e.mail address in the mailing list in order to make him =20
> aware of this discussion.
>
>
>

HyperCuP as such as used by us to handle efficient communication =20
between peer libraries. There are two services we use it for - =20
distributed search and distributed authentication and authorization. =20
With the upcoming support for OAI-PMH (finally!) and improved MMS, we =20=

are getting closer to realizing my vision of heterogeneous networks =20
of digital libraries. It would be nice however, to have some other =20
parties working with us in order to test the ideas with their =20
solutions, i.e., other libraries. The OAI-PMH support should be =20
delivered by mid-Nov the latest.
We will keep you posted on the jeromedl-users mailing list about it.

Cheers,

S.

>
>
> -----Messaggio originale-----
> Da: Sebastian Kruk [mailto:seb...@de...]
> Inviato: venerd=C3=AC 21 settembre 2007 23.19
> A: Giuseppe Bux
> Cc: JeromeDL Users; Devel JeromeDL; Maciej D=C4=85browski; hypercup-=20=

> de...@li...; Slawek Grzonkowski
> Oggetto: Re: Towards a more deep investigation of JeromeDL !
>
> X-Mailtags: =20
> YnBsaXN0MDDUAQIDBAUGBwpZJGFyY2hpdmVyWCR2ZXJzaW9uVCR0b3BYJG9iamVjdHNfEA=20=

> 9OU0tleWVkQXJjaGl2ZXISAAGGoNEICVRyb290gAGvEBgLDB0eHyAhIigpKisyMzQ1PUNE=20=

> R0tOT1NVJG51bGzTDQ4PEBYcWk5TLm9iamVjdHNXTlMua2V5c1YkY2xhc3OlERITFBWAB4=20=

> AMgA2ADoAPpRcYGRobgAKAA4AEgAWABoAXXxAQbWFpbFRhZ3NLZXlXb3Jkc18QFG1haWxU=20=

> YWdzRHVlRGF0ZVN0YXRlXl9fcHJvcG9zZWRUYWdzXxAPbWFpbFRhZ3NQcm9qZWN0XxAPbW=20=

> FpbFRhZ3NEdWVEYXRl0g0PIyejJCUmgAiACYAKgAtZQEZvbGxvd3VwWGplcm9tZWRsWGh5=20=

> cGVyY3Vw0iwtLi9YJGNsYXNzZXNaJGNsYXNzbmFtZaMvMDFeTlNNdXRhYmxlQXJyYXlXTl=20=

> NBcnJheVhOU09iamVjdBABXxAkRTJBN0I5RDgtNjRBNi00ODNELTlBRDctQUJCQTVBMDEz=20=

> Mjk1WEplcm9tZURM1DYPNzg5Ojs8WU5TLmZvcm1hdFdOUy50aW1lW05TLnRpbWV6b25lgB=20=

> WAFiNBqVtFpgAAAIAQ0z4/=20
> D0BBQldOUy5kYXRhV05TLm5hbWWAEoARgBRdRXVyb3BlL0R1YmxpbtI=20
> +D0VGTxEFBVRaaWYAAAAAAAAAAAAAAAAAAAAAAAAACgAAAAoAAAAAAAAA5QAAAAoAAAAQm=20=

> yazkZvWCxGczzCgnaTDoJ6cnaCflxqgoIW6IKF2/KCiZZwgo3vIoKROuKClP/sgpZQ/=20
> AKYlYCCnJ8YgqCosIKjr=20
> +KCqANOgqtUVIKvp8CCsx2wgrcnSIK6nTiCvoHmgsIcwILGS0KCycEygs3KyoLRQLqC1SV=20=

> ogtjAQoLcydqC4D/KguRJYoLnv1KC66QAgu9jxILzbVyC9uNMgvrH+oL=20
> +YtSDAmxsgwXiXIMJ6/=20
> SDDWHkgxFGkoMU4WyDGOsEgx1jWoMfaCaDUSdIQ1R4hoNZOnhDXLCgg2C6OINj5lSDaDnA=20=

> g2uvsINvlF6Dcy84g3cT5oN606qDfrhYg4JTMoOFySKDia3Qg41IqoORUkKDlMgyg5j2tI=20=

> OcbKSDoFFSg6PsLIOn9cSDq2u0g691TIOy6zyDts/qg7pqxIO+BZ6Dwn30g8WFJoPJ/=20
> XyDzSmYg9F9BIPUhDaD2PyMg9wDvoPgfBSD44NGg+f7nIPrAs6D76AOg/=20
> HuroP3Hu3ADcMYgBClYIAVQqCAGCTogBzCKIAfpHCAJEGwgCcj=20
> +IArwTiALshqgDNAwIA2R/=20
> KAOsBIgD3HeoBCZLqARUcCgEnkQoBMxoqAUWPKgFSPrkBY4xpAXA82QGBiokBjjr5AZ=20
> +IqQGsORkBvhpxAcrK4QHcGJEB6MkBAfoWsQIGxyECGBTRAiTFQQI2EvECQsNhAlSkuQJg=20=

> wYECcqLZAn9TSQKQoPkCnVFpAq6fGQK7T4kCzJ05AtlNqQLqm1kC90vJAwiZeQMOckADFd=20=

> 2RAycrQQMz27EDRSlhA1HZ0QNjJ4EDb9fxA4G5SQON1hEDn7dpA6vUMQO9tYkDymX5A9uz=20=

> qQPoZBkD+bHJBAZiOQQYQ5EEJGBZBDZBsQRCXnkEVD/=20
> RBGBcmQRyPfEEfu5hBJA8EQSc7IEErjoxBLrqoQTMy/=20
> kE2OjBBOrKGQT25uEFCMg5BRV4qQUmxlkFM3bJBUTEeQVRdOkFYsKZBW9zCQWBVGEFjXEp=20=

> BZ9SgQWrb0kFvVChBcoBEQXbTsEF5/8xBflM4QYF/=20
> VEGF96pBiP7cQY13MkGQfmRBlPa6QZgi1kGcdkJBn6JeQaP1ykGnIeZBq3VSQa6hbkGzGc=20=

> RBtiD2QbqZTEG9oH5BwhjUQcVE8EHJmFxBzMR4QdEX5EHURABB2LxWQdvDiEHgO95B40MQ=20=

> Qee7ZkHqwphB7zruQfJnCkH2unZB+eaSQf45/=20
> kAEDAgMCAwIDAgMCAwUEAwQDBAMEAwQDBAMEAwQDBAMEAwQDBAMEAwQDBAMEAwQDBAMGBQ=20=

> YFBgMEAwQDBAMEAwQDBAMEAwQDBAMEAwQDBAMEAwQDBAMEAwQDBAMEAwQHCQQDBAMEAwQD=20=

> BAMEAwQDBAMEAwgJCAkICQgJCAkICQgJCAkICQgJCAkICQgJCAkICQUICQgJCAkICQgJCA=20=

> kICQgJCAkICQgJCAkICQgJCAkICQgJCAkICQgJCAkICQgJCAkICQgJCAkICQgJCAkICQgJ=20=

> CAkICQgJCAkICQgJCAkICQgJCAn///=20
> oPAAAAAAgfAQQAAA4QAQgAAAAAAAwAAA4QAQQAAAAAAAwAAA4QAQQAAA4QAAQAAA4QAQQA=20=

> AAAAAAxETVQASVNUAEJTVABHTVQAAAABAQEAAAABAQAAAAAAAAAAAQGAE9IsLUhJo0lKMV=20=

> 1OU011dGFibGVEYXRhVk5TRGF0YdIsLUxNok0xWk5TVGltZVpvbmVfEBQlWS0lbS0lZCAl=20=

> SDolTTolUyAletIsLVBRo1FSMV5OU0NhbGVuZGFyRGF0ZVZOU0RhdGXSLC1UVaNVVjFfEB=20=

> NOU011dGFibGVEaWN0aW9uYXJ5XE5TRGljdGlvbmFyeQAIABEAGwAkACkAMgBEAEkATABR=20=

> AFMAbgB0AHsAhgCOAJUAmwCdAJ8AoQCjAKUAqwCtAK8AsQCzALUAtwDKAOEA8AECARQBGQ=20=

> EdAR8BIQEjASUBLwE4AUEBRgFPAVoBXgFtAXUBfgGAAacBsAG5AcMBywHXAdkB2wHkAeYB=20=

> 7QH1Af0B/wIBAgMCEQIWBx8HIQcmByoHOAc/=20
> B0QHRwdSB2kHbgdyB4EHiAeNB5EHpwAAAAAAAAIBAAAAAAAAAFcAAAAAAAAAAAAAAAAAAA=20=

> e0
> X-Mailer: Apple Mail (2.752.3)
> Return-Path: seb...@de...
> X-OriginalArrivalTime: 24 Sep 2007 08:40:12.0799 (UTC) FILETIME=3D=20
> [85D8A8F0:01C7FE86]
>
> -----BEGIN PGP SIGNED MESSAGE-----
> Hash: SHA1
>
> Dear Giuseppe,
>
>
> On 13 Sep 2007, at 11:38, Giuseppe Bux wrote:
>
> >
> > Dear Sebastian,
> > As I let you known, yesterday I made a demo of JeromeDL to my
> > project team
> > (dContentWare project): it resulted interesting for the whole team,
> > so we
> > decided to more deeply investigate the system functionalities and
> > customization capabilities, as well the L2L communication over the
> > http ELP
> > protocol.
> > As to the system functionalities, on the 2.0.2.jar  release I do
> > not have
> > the running troubles I had on the 2.0.1 one, and that I informed
> > you about
> > in my previous email. However some not effective functional issues
> > still
> > remain open as follows:
> >
> > 1) Resource uploading: the selection of the XSL:FO format do not
> > display any
> > uploading resource form:what about its usage?
>
> As I have already mentioned in the other email - XSL:FO is the
> feature delivered by GUT. It was used to hold info from RTF-to-XSL
> transformation (using RTF2FO service) and used to render HTML, PDF,
> and RTF versions of a resources.
> Since we are short on resources at the moment in DERI, I have already
> asked GUT if they could update XSL:FO support in recent versions of
> JeromeDL.
>
>
> > 2) Search for resources:
> >       2.1)Advanced search:
> >               2.1.1) it works only with parameters provided by =20
> jeromeDL
> > metadata scheme;
> > I assume that in next JeromeDL                                =20
> release it will work also with
> > parameters
> > extracted from the other metadata scheme available in the =20
> pop                         up
> > window;
>
> actually this is very good idea - I will put it on my todo list :)
>
>
> >               2.1.2)  How to set  the searching by using Domain =20
> and Keyword
> > tagging
> > values? How to gain from the semantic                         =20
> nature of tag values
> > extracted
> > either from Wordnet or from Taxonomies?
>
> with current stable versions of JeromeDL - you cannot do too much,
> but upcoming 2.1 version (currently beta stage - see
> bleedingedge.jeromedl.org) this information can be used in the
> faceted navigation view - MultiBeeBrowse. We also continue working on
> JOnto component - to provide more support for Wordnet and taxonomies.
>
>
> >
> >       2.2) Semantic search - Natural Language Query-RDF Query:
> >               2.2.1) it proposes a limited set of templates for =20
> RDF queries; I
> > think
> > that a more general RDF template                              =20
> providing the user with
> > couples of
> > metadata-values  (such as the one adopted for advanced =20
> search)                                and
> > related concerned constraints should be provided;
>
> Please consider it a proof of concept. I am very interested in
> finding any actual use case scenarios for these NLQ component. If you
> have any questions in mind - let me know - I will try to deliver them.
>
> >               2.2.2); templates for RDF search concerned with =20
> domain or keyword
> > values,
> > possibly managed as semantic                                  =20
> concepts, are necessary.
>
> ok, I hear you - I will try to cover these questions in the next 2
> weeks. Will keep you posted.
>
>
> >               2.2.3) What about the search based on NLP algorithms?
>
> I do not have enough expertise, and I have no resources to
> investigate it further at the moment.
> If your institute has some resources you would like to contribute to
> this feature, I will be happy to assist you.
>
> > 3) Resource rendering:
> >               3.1) Rendering of resources: rendering resources =20
> under print
> > protection
> > enacts an applet for image                                    =20
> displaying  that remains in a
> > started status
> > without any image display.
> >
>
> Hmm, let me investigate it.
>
> > 4) Metadata mediation services:
> >               4.1)The mapping from an  RDF resource description =20
> to another
> > metadata
> > scheme (ex. Dublin Core, or BibTex, etc.)                     do  =20=

> not concern the whole
> > resource description but only the resource identification metadata;
> > this is
> > a                     strong  limitation in a Digital Library's =20
> communication
> > environment
> > (see below).
>
> JeromeDL uses external service called MarcOnt Mediation Services -
> Maciej is working on this services at the moment.
> So the quality of mappings is to improve over the time.
>
> >
> > 4)Hypercup support: Where documentation about? How to set the query
> > web
> > service?
> >
>
> The hypercup is handled by Slawek.
> I have copied him on this email.
> He will assist you.
>
> Cheers,
>
> S.
>
>
> >
> > looking for your kind answers,
> > Best regards,
> > Giuseppe
> >
> >
>
> - --------------------------------------------
> - --   Sebastian Ryszard Kruk
> - --   Lead Researcher, Project Manager
> - --   Semantic Infrastructure Lab, eLearning Cluster
> - --   Digital Enterprise Research Institute
> - --   National University of Ireland, Galway
> - --   mailto: seb...@de...
> - --   GG: 335067, Jabber: s_...@ch...
> - --   Skype: sebastiankruk
> - --   WWW: http://www.sebastiankruk.com/
> - --   mobile (IRL): +353 85 7126591
> - --   VoIP   (PL):  +48  52 5110114
> - --------------------------------------------
>
>
> -----BEGIN PGP SIGNATURE-----
> Version: GnuPG v1.4.7 (Darwin)
>
> iD8DBQFG9DUx24FHaIBwxVIRAs2CAJ93yfjQ0Z7jubbkm1kZ3ZbpJ6peAgCcDdzf
> UFdDFoAvcWrrwKbZMF6p5k4=3D
> =3DlspI
> -----END PGP SIGNATURE-----
>
>

- --------------------------------------------
- --   Sebastian Ryszard Kruk
- --   Lead Researcher, Project Manager
- --   Semantic Infrastructure Lab, eLearning Cluster
- --   Digital Enterprise Research Institute=20
- --   National University of Ireland, Galway=20
- --   mailto: seb...@de...
- --   GG: 335067, Jabber: s_...@ch...
- --   Skype: sebastiankruk
- --   WWW: http://www.sebastiankruk.com/=20
- --   mobile (IRL): +353 85 7126591
- --   VoIP   (PL):  +48  52 5110114
- --------------------------------------------


-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.7 (Darwin)

iD8DBQFHBezj24FHaIBwxVIRAjthAJ0cXo1xfVL6CJWWQDgj9hV4FkpV7ACgsyIu
HqLD263KH3wmp8nycHLXqvA=3D
=3DzUuH
-----END PGP SIGNATURE-----