formism

This is a side project we are investigating with some friends in order to see if contemporary machine learning methods such as word2vec can be used to produce social mapping with layers of complexity.

Distributed computing over core grids is also a playground to look at viability of reading maps based on networks. After a simple statement formalization it is easy to see if hypothesis are strong over ten of thousands of different maps produced with the same corpus and samples of that corpus.

• python notebooks

cortext manager

The goal of the cortext manager platform is to deliver a ready-to-go tool for humanities and social sciences researchers to perform text analysis over big corpora. Beside integrated text mining algorithms, the main up value was the integration of network visualization and mapping.

In this context, my task was to drive the design and development of the plaform through service design analysis and information architecture.

• cortext team github repository

The project

Contacted by the UN Global Pulse initiative, we were part of research dream team composed of scientits affiliated to ISC-PIF and IFRIS.

We have then been asked to build a prototype using real-time data to monitor world-wide issue of our choice. By common concerns, we chose to work on the problem of media framing of the food security public issue.

For this project, we made an heavy use of concepts maps as blueprints for the organization between content and interaction. As functionnalists, we primed minimalism between the use and the form.

Maps were great tools to communicate and get quick consensus on fundamental understandings of the team.

Information architecture was important as a set of skills in order to share complex concepts, various point of views and transforming the collective works as a sustainable space of ideas and concepts.

Even if it was requested by the brief, we tried to push the innovative concept of web of objects to a certain reality. We then provided a full transversal framework of independant information designed dispositifs. The website is used as intended but is also augmented by the eco-system in which it is taking part.

The idea is not new. We just take it as something to realize. Most of the concept and analysis work can be found in books such Bruce Sterling's Shaping Things.

The project

IRIS was a data visualization project commissioned by heaven. The main purpose of the project was to show to clients the effectivness of the community management teams.

It was then important to have a scalable visualization that permit comparison between periods.

IRIS can also be used as representation of differential amount of work between clients and missions.

MultipliCity Disorder was a collaboration with Angelique Chemois (architect) and Ola Wasilkowska (architect) and a proposal to the Europan 9 contest. We tried to invest into network theory and ecosystem modelling in order to open up a new way to think the territory and the relationship between the architects, urban planning and the communities.

The main idea was to use a prey-predatory model to constraint diversity within a landscape. We also played with the concept of strangification to develop the idea that architectural intervention needed a translation effort.