StreamStory is a tool, which which can extract a structure and regularities within the data. It can show the state of the monitored process, activity and could be also used as anomaly detection tool. StreamStory could be accessed on http://streamstory.ijs.si.

StreamStory can help you interpret temporal data of different kinds. It is useful for the analysis and exploration of large multivariate time series. StreamStory has several mechanisms to uncover and in particular explain the structure within the data. These mechanisms are visual (hierarchical Markov chain, charts, decision trees, parallel coordinates) and also a textual narrative explaining/summarizing states and patterns within the data.

StreamStory uses sparse feature vectors as its input. with our tool you can transform the dynamic network into a sparse feature vectors. Input data for a StremStory tool could be in numeric or categorical form, and you can choose to export data in strictly numeric form or with text labels. Using labels is recommended, since StreamStory tool can use these labels as a textual narrative explaining or summarizing states and patterns within the data. This makes users to much easily uncover the dynamic of the public spending and to spot regularities and anomalies in public spending data.

Through the web interface, you can select public procurement data for one or more contracting authorities or one or more economic operators. Also, a group of specific contracting authorities and economic operators (who had the “relationship”, i. e. economic operators who won the contracts for specific contracting authorities) could be selected.