Aspects of Information Integration in a Wiki

Information integration means pulling information pieces from various sources together with as little loss, as little overlap and as little redundancy of information as is possible. This paper explores different aspects of information integration using as an example a wiki that contains a set of documents to which we want to add suitable further pictures. We take into account a number of important aspects such as: (i) Pictures outside the domain of the wiki are ignored. (ii) Pictures that already occur in the wiki or very similar to existing ones are also ignored. (iii) Each new picture has to be associated with a suitable document in the wiki subject to three restraints: if a suitable document does not exist, a new one is created; if the document size due to pictures gets too large some additional actions are necessary; an image may be associated with more than one document. Each of the above points offer many challenges which are discussed.