Expert Finder is an application for ontology to find the best possible expert from the user profiles in the ontology for the search field entered which is either a course, area of research or even a project related to a research field. Retrieving instances from ontology of an expert finder is considered as a problem and in this thesis the different possible ways to overcome this problem are analysed and evaluated. The instances in the ontology are retrieved and ranked theoretically to provide the most qualified expert for the query, which will be done by using various available methods to retrieve the instances and rank them. The results and the procedures used for finding the expert are compared with each other to analyse their advantages and disadvantages. Although there are various possible ways using distributed computing and using databases, it was the aim to work only on ontology rather than integrating various available methods which generally complicates the task. Distance mapping and Agglomerative hierarchical clustering are two methods which are analysed in depth and tailored to the situation to provide the required result from the expert finder ontology on which these methods are implemented theoretically. Depending on the situation of the problem, some of these methods are used partially and tailored according to the suitability to the problem that is being analysed.