ABOUT

Nifty.works is a personal project focused on information retrieval.
Nifty.works is a web service to visualize the context of a thing and speed up learning and information retrieval.

Alpha version has been shown at annual conference of KnowEscape 2015, a research and development action funded by Horizon 2020 European Union Program.

Approach:
1. Benchmark with current javascript libraries (d3.js, vivagraph.js, sigma.js)
2. Extending libraries functionalities and tuning performance: achieving responsive interaction with connected graphs in SVG, 300+ nodes
3. Architectural information full-stack, benchmark and testing with graph databases and noSQL database with python data-manipulation in real time; Development of API REST; Python Flask + ElasticSearch
Implemented graph algorithms to find out logical paths between two nodes
4. Material design
5. Restyling wikipedia content with material design and hijack links for consistent visualization
6. Enhancing Wikipedia with discovery functionality

UPDATE July 2016
- Query for context
implemented full-connected graph display
- Implemented query for suggestions shared between topics!
You can know query for pathways between topics and speed up your brainstorming.
As example, what's in between 'unemployment' and 'financial crisis'?
Or, 'dna' and 'gene editing' and 'neuroscience'?
Go and explore.

EDIT AND SHARE MAPS
Data visualisation of connected graph can be edited and shared.
Currently I find out a bug on Facebook for which title of object is not properly set (see bug number 487310988131895 on developer.facebook site). Waiting for a fix :)

ROLES
For this project I covered:
- Ideation
- Design
- R&D on Knowledge graphs, recommender systems, network science
- software architecture: back-end, front-end
- software development
- UX increased readibility graph-layout of 40% for expanded on demand graph layout
- UI responsive mobile