In order to satisfy the requirements determined by the given types of users, the INFRAWEBS Conceptual Model (Fig. 1) has been designed. It is based on the basic approaches of the INFRAWEWBS framework and is a conclusive evolution of the formerly defined global design.

Fig. 1: The INFRAWEBS Conceptual Model

The project work takes into consideration different dimensions and connects comprehensively the different INFRAWEBS components. The INFRAWEBS Conceptual Model consists of two main elements: the INFRAWEBS Semantic Web Unit and INFRAWEBS Environment.

The INFRAWEBS Semantic Web Unit provides users with:

• Information structures for effective discovering, storing and retrieving both semantic and non-semantic information needed for creating and maintaining Semantic Web Services.
  
  
• Tools for creating and maintaining Semantic Web Services and SWS Applications.
  
  
• Library of Methods used for creating and maintaining Semantic Web Services.
  

The INFRAWEBS Environment provides means for communicating with different kinds of INFRAWEBS users and other INFRAWEBS Semantic Web Units as well as for executing SWS considering security and privacy issues of these operations.
All SWU components are organized in two directions:

• Problem solving based on semantic information (or Logic-based problem solving) versus problem solving based on non-semantic information (similarity-based problem solving).
  
  
• Different types of information needed for both kinds of problem solving.
  

From this point of view the SWU proposes:

• Information structures for storing and retrieving semantic and non-semantic data that are DSWS-R (Repository) and Similarity-based OM.
  
  
• DSWS-R is aimed at effective storing and retrieving all elements of Semantic Web Services according to the WSMO Framework: Goals, Ontologies, SWS and Mediators written on WSML language.
  
  
• a Similarity-based OM contains non-logical representation of the same (or some of the) elements (so called "knowledge objects") as well as other data e.g. graphical models of SWS and "natural language" templates for WSMO Goals.
  

Tools for creating and maintaining both semantic and non-semantic data:

• CBR based Service Designer is aimed at designing a WSMO-based Semantic Web service from an existing non-semantic Web service.
  
  
• CBR based Service Composer is aimed at creating a Semantic Web service through composition of existing WSMO-based semantic Web services.
  
  
• CBR based Goal Editor is aimed at creating predefined WSMO-based goals and their "natural language" templates needed for designing SWS-based applications.
  
  
• CBR based Recommender tool is a similarity-based tool facilitating operation of all INFRAWEBS "semantic-based" tools by utilizing "past experience". Non-semantic data stored in OM is used as the problem description for determining the most similar solution (SWS or its graphical model) to the current problem.
  
  

Methods used during problem-solving

• Logic-based discovery
  
  
• Application-specific decision-support methods used for service composition, compensation, monitoring etc.
  
  
• Ontology keywords based discovery
  
  
• Several methods for calculating similarity and/or assignments - structural, linguistic, statistic etc.
  
  

Semantic and non-semantic components of INFRAWEBS are interconnected by the Semantic Information Router (SIR), which is responsible for:

• Locating all resources needed for problem solving either in the local SWU or outside.
  
  
• Creation of non-semantic content (knowledge objects) by means of the semantic content stored in the DSWS-R.
  
  
• Creating an effective system of indexes allowing fast communication between semantic and non-semantic modules of the SWU.