This article deals with polynomial algorithms for community detection in networks which satisfy some useful interpretation when applied. The basic theory necessary for the understanding of community detection algorithms is explained, as well as some community detection algorithms and properties. The algorithms described are part of a concept called hierarchical clustering. The main focus of hierarchical clustering is transforming of a vertex partition into a new vertex partition, thus creating a hierarchy. There are two approaches, one of which is agglomerative, and the other divisive. The agglomerative approach attempts to increase the number of partitions by one at each iteration, whereas the divisive approach attempts to do the opposite, decreasing the number of partitions by one at each iteration. We have described the best known examples of algorithms for each approach – the Ravasz and the Girvan-Newman algorithm, respectively. Apart from that, we have described a greedy algorithm for optimizing modularity which can be implemented in both approaches. The result of hierarchical clustering is a family of community series, which we have visualized by using dendrograms. The efficiency of the non-optimal, but equivalent Python implementation of the algorithms decribed above, as well as the quality of the results measured by modularity has been conducted on randomly generated networks, with the help of a Python library called NetworkX.