| A. What is Papyrus about? |
| |
Papyrus is about a Digital Library created dynamically on demand of a user query for content. |
| |
| B. What is so special about this Digital Library? |
| |
The unique feature of Papyrus is that it will understand the user query in the context of a specific domain (e.g. history) and, while drawing content from a different domain (e.g. news), will present the results in the context of the domain of the query (i.e. history). |
| |
| C. So, is this a search engine? |
| |
No. A search engine neither understands the query in a specific context nor does it organize the results as a library. Papyrus introduces the concept of a semantic digital library engine. |
| |
| D. And how could news be presented as history? |
| |
Although both news and history are about events, each domain follows its own model. For example, a news item conveys the view of a journalist on a specific event: when it happened; how it happened; who were involved. History is about why it happened, what the consequences were and how it could have been avoided. Papyrus will model both domains and will try to map the correspondences.
|
| |
| E. Will complete models of news and history be created from scratch? |
| |
No. Existing standards experienced by the consortium partners (e.g. the IPTC news categorization) will be extended in depth in specific sub-domains (e.g. history of science and technology) to prove the viability of the concept. |
| |
| F. So will Papyrus be just about mapping news to history? |
| |
No. This will be the illustrated use case. The technologies developed will target the generic case of building dynamic digital libraries across domains that share common characteristics. Another potential use case could be built around law and the history of politics. |
| |
| G. What will the necessary research be about? |
| |
Advances will have to be achieved in the areas of context based query analysis, knowledge mapping and multimedia content analysis. |
| |
| H. So will Papyrus perform automatic multimedia content analysis to index the source content? |
| |
Professional news content is already annotated, although in an unstructured and inconsistent way. Papyrus will exploit these annotations to perform a targeted content analysis. Automatic multimedia analysis is impossible to realistically hope for without restricting the problem to a very narrow domain. |
| |
| I. So will Papyrus finally be a portal? |
| |
The target is to build Papyrus as a web based system so that distributed source archives can be easily connected to provide for a scalable implementation. |
| |
| J. Does the consortium have the expertise to make this happen? |
| |
The Papyrus consortium includes 4 research organizations with specific expertise in the scientific areas addressed, 2 research organizations experts in the history of science, 2 worldwide leaders in news and broadcasting, the industry world leader in the market of search engines and an SME with relevant products and a long track record in coordination of similar projects.
|
| |