Delight in Thinking

L inked data can be the basis for a new generation of Semantic Web has a very broad application prospects. W3C recently announced, SKOS Linked Data will be full of. (Needle, 2009) which is significant for the library community. SKOS of the linked data should benefit from the libraries of the United States Congress, the Library of Congress Subject Headings (Library of Congress Subject Headings) of all the Linked Data, Linked Data application to become a successful example, promoted the Linked Data to practical. This in turn can be said that libraries contribute to the development of the Internet instance.

L inked Data are based on RDF and http protocols, in other words, Linked Data is to pass through the Http RDF data, and the RDF data is a special rule to follow the semantics of data, this rule is that any resources are used to characterize the URI. So URI \ RDF \ HHTP constitute the three cornerstones of Linked Data. The application of any Linked Data revolve around these three aspects of the. Such as:

• How to build a resource URI?
• How to create RDF resources?
• How to use http to transfer RDF?
• How to accept and parse RDF data resources? And so on.

L inked Data Application There are two basic models: a model for the browser / server mode, that people through the browser to access Linked Data; The other model is a client / server model, where clients mainly refers to the application system, the model is actually the system and data transfer between systems. Whatever the mode, Linked Data is to pass through Http.

H ttp protocol actually defines two processes, one request, the other is the response. To these two processes as the blueprint, we can simulate a Linked Data transfer process.

• Client-side or browser sends a request to the RDF file server, requests the server to return an RDF document. Send this request is simple, you can directly enter a rdf file suffix, such as http://www.cloudlibrary.info/rainzenfoaf.rdf rdf is a simple request. Of course, this request is very weak, the file suffix is not necessarily with the rdf for RDF format, in turn, RDF format file suffix it may not rdf. So, to pass with the file extension is not very reliable RDF request. What we need is a mandatory request RDF very clear way, this request can only be through the system (such as the browser) rather than be issued manually. Now a common browser can not issue such a request, in order to apply Linked Data RDF we need to send a request to the browser.
• RDF when the server receives a request, need to return an RDF document, and inform the client that this is a RDF file. This process is called content negotiation. To achieve this RDF-based content negotiation, which would set the MIMEType, the definition of an RDF type. RDF 2004, one on the content type: Application / RDF + XML to be registered. In terms of the Linked Data, content negotiation is very important, if you want to publish Linked Data, require the server can implement content negotiation.
• Device sent a service response to a client, particularly the browser, how to parse the RDF file? Particularly in the browser mode, because RDF is not for seeing, but to the machine to see how the machine language translation of the expression of adults to read? This in turn requires a special browser to parse RDF.

View from the process, applications linked data needs:

• RDF can send a request, and be able to resolve RDF browser, we call RDF browser. Keven described in his blog a lot of RDF browser or plug-ins. (Wei, 2009)
• A server to return RDF type of server. Fortunately, we often use the Apache open source http server: that is one of the few to achieve a good content negotiation server.

Published in the Linked Data using Apache, you need a simple set of the Apache needs to be added regarding the RDF of content-type, is set in Apache, or in. Htaccess add a type description: AddType application / rdf + xml rdf . This Apache receives a request for RDF, RDF can be sent back to type messages.

However, as application / rdf + xml is the new register, and some browsers do not compatible, I found a phenomenon, visit http://www.cloudlibrary.info/rdf/rainzenfoaf.rdf when the software by checking the file header that, the server returned an application / rdf + xml document, but by Firefox to see the file information, been interpreted as text / html. It seems firefox's default content-type is text / html, and not authorized application / rdf + xml type, when receiving this type, only use the default value.

Another interesting phenomenon is that, Apache's default type is text / plain, if no RDF type, the server can not confirm RDF file to start the default value, the RDF file as text / plain. Return the text file information, then the browser displays the text file. So, is the need to increase application / rdf + xml then this type? Is not necessarily a matter of fact, already stated, FireFox the application / rdf + xml interpreted as text / html, then we add a type: AddType text / html rdf also line, or directly to the Apache's default value is set to text / html on the line. So how do you explain the return to send over the text / html message, the browser was able to resolve it into RDF? It seems that the browser client also needs a judge, by analyzing the markup language to determine whether the RDF file. However, in the communication between systems, I think, or should be enabled application / rdf + xml, so clearly more reliable.

This text is formed during the discussion and Keven, thanks to Keven contribution to this article.

References:

Needle, David. 2009. W3C's New SKOS Standard Advances Linked Data. [Online] August 18, 2009. [Reference date: August 27, 2009.] http://www.internetnews.com/dev-news/ article.php/3835176 .

Wei, Liu keven. 2009. Associated data browser. Number Chart Research Notes. [Online] August 26, 2009. [Reference date: August 27, 2009.] http://www.kevenlw.name/archives / 1844 .

Background Links

http://www.w3.org/Protocols/

http://www.iana.org/assignments/media-types/application/

LIS research is being reborn, the recently published draft FRSAD again confirms this change. And FRBR, like, FRSAD is a concept model, is used to describe the basic framework of human understanding of the world, and traditional library science different, FRSAD the study made use of a more precise formulation of the structure, making their research more is easy to be resolved, coding and processing. When people discuss the Library of precision, FRSAD demonstrated a feasible approach.

F RSAD used to describe the core of our understanding of the world, FRSAD the draft to explain their purpose in order to FRBR work (work) in the "on" (aboutness), a conceptual model. When we say "This book is about the 'Late Qing China, the parliamentary system' and" What is the meaning of this sentence? How to formally express the meaning of this sentence? The logic of this statement what is the structure? FRSAD trying to explain the program gives a. Obviously, FRSAD and FRBR is closely linked to her own claims of Group 3 for the model FRBR, ((FRSAR), 2009) but I personally think that, FRSAD is significant not only in this, but more importantly, a more a wide range of cognitive models constructed to describe the world, in other words, the whole "objective knowledge" in the world to model.

Because FRSAD intentionally or unintentionally, the model for objective knowledge, then the starting point is very FRSAD work of philosophy, to a large extent, FRSAD can be seen as an epistemological model, if the back FRSAD philosophical roots can be traced about to 亚利斯多德 time ontology.

F RSAD is the core of Thema-Nomen model, this model assumes that any work (work) [Note] have a theme, the theme is called Thema (CIMA), any CIMA can be expressed as per Nomen (Norman). ((FRSAR), 2009) corresponds to the work (Work), CIMA is also the abstract, but Norman was the objective expression of CIMA, which in various languages, symbol tables Daxi Ma. This assumption is very important that this is the translation of any subject, indexing, authority control basis. Late Qing China held a parliamentary system, with regard to the subject can be expressed as keywords in Chinese, English can also be expressed as keywords, can also be expressed as the Chinese library classification number, can also be expressed as the Library of Congress Key words. Be as many as I understand it should belong to Norman area in which CIMA is only one. Because the presence of CIMA, the library can be a different indexing system can be used to describe the same document, only between different indexing systems may each map together.

T hema-Nomen model explained the main contents of Thema and Nomen attributes and relationships, which define the Thema-momen relationships between the key to the model, these relationships include

• CIMA and the relationship between Norman;
• The relationship between CIMA and CIMA;
• Norman and Norman the relationship between

F RSAD has been defined in detail relationships, see text.

In practice, the difficulty is the definition of various per class CIMA, you can use these dollars into the Five Elements class definition, it can also be used Ruan Gang praise in the sub-surface classification of the various surfaces.

F RASD is a huge system, not one or two words to say clearly, nor is it a day or two to fully understand. The discussion above is only some of my preliminary view, may be not very serious misunderstanding, right when initiate a more detailed review will be published in the near future in his new blog http://www.cloudlibrary.info on.

In addition, FRASD project team is chaired by all of us are familiar with Professor Zeng Lei, so we will feel when reading FRASD cordial.

References:

(FRSAR) Working Group on Functional Requirements for Subject Authority RecordsIFLA. 2009. Functional Requirements for Subject Authority Data (FRSAD): A Conceptual Model. Draft, 2009 年. [Reference date: June 2009 on 29th day.] Http:// nkos.slis.kent.edu/FRSAR/report090623.pdf.