Rezultaty
The most outstanding elements of eStage's human-computer interaction are concept hierarchies that always provide the users with the necessary orientation while exploring the information space. Unlike in traditional search engines, eStage users search or navigate only by selecting entries (terms or phrases) in concept hierarchies. Defining a query means simply clicking on entries. Each of these hierarchies, in turn, stands for a superior aspect or for a category. For instance, the information space can be grouped according to categories like productions (spectacles), attributes, and personalities. In such an information space the user never loses the semantic context or encounters data that belongs obviously to a completely different context.
Grouping large data samples along easily discernible information spaces tackles the problem of ambiguity that the traditional search engines cannot master. This means that at each navigation step a user gets only a small set of documents, precisely tailored to the selected entries in the respective hierarchy. The interface and the concept hierarchies are available in four different languages (English, French, Spanish, and German). The architecture of the underlying client-server system in eStage completely resorts to the design developed in IRAIA. The main design objective for both systems is to catch with a rapid growing process of web technology.
A too rigidly designed platform suddenly turns into something impossible to change or create new applications with months of work and constant crashes. We therefore developed a system of independent layers. The communication protocols between the layers ensure their seamless integration. The browser client is realised by a set of Applets and Java-scripts. We therefore redistributed some parts of the client component and the communication processes among Java-servlets and Java-script routines.
On the backside of the system we still resort to the demon existing from the beginning of the IRAIA project with its approved and stable SQL querying routines. eStage handles multimedia documents that are mapped into concept hierarchies (CHs) for organisation and access purposes. These associations can be done manually or automatically on the base of the textual content of the documents. The association of textual documents to CH corresponds to a categorisation task. Most of the applications in text categorisation aim at deciding which group or class a document belongs to. However, the assumption that a document can be classified in different categories is more realistic than considering a document can be clearly identified as belonging to a unique category.
In eStage, like in IRAIA, a category corresponds to a node from a CH. A document can be categorised in different nodes from a CH, in addition, it can be categorised in several CHs according to its contents. This is an important feature linked to the use of CHs to query the document collections. Browsing the CHs leads to the selection of relevant concepts and the definition of a query composed of selected concepts and their description. Expression of a query becomes an instantiation of several concepts.
The association of textual documents to CHs is based on a "vector voting method". This method is based on a voting mechanism. The basis of the process is that the more terms from a category occur in a text, the stronger is the link between the text and that concept. In fact, the principle is more complex as the terms corresponding to a category are not only terms from the concept itself, but also those associated with the category ancestors and semantically linked terms (eventually resulting from a learning phase). I
Wyszukiwanie danych OpenAIRE...
Podczas wyszukiwania danych OpenAIRE wystąpił błąd
Brak wyników