Clustering is an unsupervised learning method that determines partitions and (possibly) prototypes from pattern sets. Sets of numerical patterns can be clustered by alternating optimization (AO) of clustering objective functions or by alternating cluster estimation (ACE). Sets of non-numerical patterns can often be represented numerically by (pairwise) relations. These relational data sets can be clustered by relational AO and by relational ACE (RACE). We consider two kinds of non-numerical patterns provided by the World Wide Web: document contents such as the text parts of web pages, and sequences of web pages visited by particular users, so-called web logs. The analysis of document contents is often called web content mining, and the analysis of log files with web page sequences is called web log mining. For both non-numerical pattern types (text and web page sequences) relational data sets can be automatically generated using the Levenshtein (edit) distance or using graph distances. The prototypes found for text data can be interpreted as keywords that serve for document classification and automatic archiving. The prototypes found for web page sequences can be interpreted as prototypical click streams that indicate typical user interests, and therefore serve as a basis for web content and web structure management.