Skip to content


PICI (Peer Innovation Community Indicators) is an open source toolbox that simplifies creating and measuring indicators of innovation in online communities. It is intended for use in a research context and allows researchers to:

  • automate data collection and indicator generation steps
  • simplify the implementation of new indicators
  • simplify the assessment of new communities with an arbitrary "digital footprint"
  • make the generation and evaluation of indicators reproducible
  • provide a repository of documented & evaluated indicators of innovation in online communities

The toolbox is a result of the Peer Innovation research project by the Chair of Innovation Economics at Technical University Berlin in collaboration with the Institute for Ecological Economy Research (IÖW). It was financed by the German Federal Ministry of Education and Research.


PICI is a Python package that requires Python \(\geq\) 3.8. Its dependencies are managed with Poetry. To install, clone the git repository and then install the dependencies with Poetry:

git clone
cd pici
poetry install


PICI is intended to use in Jupyter Notebooks (see the examples). Most toolbox features are made available through the Pici class:

from pici import Pici

p = Pici()

On instantiation, PICI collects data on the specified communities or loads them from cache. It can then be used to generate pre-defined indicators or to set up new indicators. PICI can represent indicator generation as a scikit-learn pipeline. This allows the usage of standard ML tools for further tasks, such as classification or indicator evaluation.

pipe = p.pipelines.topics()

This documentation provides further details on other aspects, such as setting up new communities or using labeled community-data for supervised learning.

If you are using PICI for research, please let us know and/or cite the following publication: (upcoming).