Clustering is one of the most significant research area in the field of data mining and considered as an important tool in the fast developing information explosion era.Clustering systems are used more and more often in text mining, especially in analyzing texts and to extracting knowledge they contain. Data are grouped into clusters in such a way that the data of the same group are similar and those in other groups are dissimilar. It aims to minimizing intra-class similarity and maximizing inter-class dissimilarity. Clustering is useful to obtain interesting patterns and structures from a large set of data. It can be applied in many areas, namely, DNA analysis, marketing studies, web documents, and classification. This paper aims to study and compare three text documents clustering, namely, k-means, k-medoids, and SOM through F-measure.

Many recent studies regarding web focused on the web services composition. The main advantage of the web services over other distributed systems technologies is the possibility of automatic services composition with previous ones so it is possible to do a series of recommended operations. Many techniques have been presented to realize the idea, but it is still challengeable how to compose such services to be both sufficient and realizable structurally. In the study we present some techniques to be used to facilitate the discovery, choice and semi-automatic composition of incongruous web services. The framework has three main qualities distinct from other activities in the same field; first the algorithm makes possible the discovery and the composition of present services according to the user's needs in low error possibility. Meanwhile, it differentiates between different levels of workflows and finally the proposed framework permits the user to distinguish and scrutinize the workflow with high authority.