The factbase is a generic dependency data table where each row describes one entity to be clustered. Each column contains the value for a specific attribute.

Agglomerative algorithms start at the bottom of the hierarchy by iteratively grouping similar entities into clusters. At each step, the two clusters that are most similar to each other are merged and the number of clusters is reduced by one.

Complete linkage gives the most cohesive clusters. Clusters formed using single linkage approach are not cohesive as complete linkage, and the results of average linkage algorithm lie somewhere between those of single and complete linkage.

Agglomerative algorithms perform the following steps:

1. Compute a similarity matrix
2. Find the two most similar clusters and join them
3. Calculate the similarity between the joined clusters and others obtaining a reduced matrix
4. Repeat from step 2 until two clusters are left