Detecting communities with the Reworc Community Detection Tool
Sometimes it is desired to see if we are able to detect communities in the population, based on the level of attraction between organizational entities, as measured with the Value Network model. For this reason, Reworcs "Community Detection" tool was created. The Community Detection can automatically create unique communities based on the Value Network model collaboration data, helped by machine learning in the platform.
How to use the Community Detection Tool
When the WorkNavigator dashboard is being used for the analysis of workstyles, an administrator accesses the "WorkNavigator Manager" to run the Community Detection:
1. You can access the WorkNavigator Manager in the top-right corner of the WorkNavigator dashboard.
2. Selects the population you want to use and expand the "People and Attributes" panel. Scroll down until you see the 'Community Detection Attribute'. Using the community detection tool will create a new attribute that will contain communities, based on level of attraction from the Value Network model collaboration data.
3. By clicking the "Create attribute using Community Detection" button the Community Detection tool is opened. A name for the attribute must be entered and an optional Prefix string can be entered that will be prefixed to each of the generated communities (groups of strong collaborating organizational entities).
4. Lastly, by clicking "Save and Close" button the Community Detection will be started and the panel will be closed. After a few sections the communities will be created. This attribute can be found in the WorkNavigator attributes menu on the right side. After this, the individual attribute values can be used in the WorkNavigator dashboard (grouping, filtering, reporting and scenario modelling) and can also be managed (renamed, reordered) in the WorkNavigator Manager.
What does the Community Detection Tool do?
In simple terms, the Community Detection tool is an innovative technology that uses the data of our value network model and applies a community detection algorithm to determine which configuration of teams in groups (communities) have the highest internal connectivity. It uses computing power and an algorithm to do this faster and more accurately than any human could do. For a more in-depth explanation, you can read the paragraph below.
The applied algorithm is a customized and proprietary algorithm from Reworc based on the Louvain method for community detection. The inspiration for this community detection method is the optimization of modularity as the algorithm progresses. Modularity is a scale value between −0.5 (non-modular clustering) and 1 (fully modular clustering) that measures the relative density of edges inside communities concerning edges outside communities. Optimizing this value theoretically results in the best possible grouping of the nodes of a given network. However, this algorithm uses heuristic algorithms because going through all possible iterations of the nodes into groups is impractical.
The Louvain community detection method finds small communities by optimizing modularity locally on all nodes. Then, it groups each small community into one node, repeating the first step. The technique is similar to the earlier method by Clauset, Newman, and Moore that connects communities whose amalgamation produces the most significant increase in modularity.