Gargantext is a web application that merges advanced tools for text-mining, network analysis and interactive visualisation; allowing new types of interaction with a given corpus. Gargantext enables its users to produce their own knowledge map or even to pinpoint the latest state of the art findings in a few minutes.
Text-mining refers to an in-depth scrutiny of texts in natural language. It processes sequences of character strings in view of obtaining a well structured set of information. It is a hybrid field that saw the light of day in the aftermath of an encounter between an array of several engineering issues awaiting to be addressed ; to cite a few of them: terms identification, themed classification of knowledge and opinions, extraction of phenomena from the news, etc.
Nowadays, modellers are building calculation codes and methods that grow more complex by the day. Understanding the behaviour of models has become an arduous and daunting task for the very researchers who have devised them, not to mention that data processing on personal computers and machines has become impossible to achieve, resulting in a catch-22 situation. OpenMOLE, a digital platform putting forward high level methods of large-scale distribution, has been particularly devised to carry on these tasks and to meet this double challenge.
«R» has been gaining prominence since its inception, 10 years ago. This (coding) language had about 30 packages ( libraries of functions) by the beginning of the 21st century (namely the years 2000s) before reaching 6000 thanks to the expansion it has undergone. Today, R allows its users to carry on statistical analyses, data manipulation, graph analysis, text analysis, spatial analysis, etc.
MULTIVAC Platform aims at helping the scientific community to overcome large-scale datasets challenges. Our existing datasets are available in a variety ranging from structured, semi-structured to unstructured. Our infrastructure streams millions of data every day ; therefore, big data architectures are designed and implemented to not only help our internal projects and team members but also to allow other scientists and researchers to take advantage of our amazing services in scale of billions of data.
Netlogo is a language that is generally used to quickly build multi-agent models. However, the Netlogo platform proposes a small number of tools to execute stochastic models simultaneously . Throughout this training, we will see how the OpenMOLE platform meets this challenge through putting forward a large set of experiments and devolving them to distributed environments without any strain.