3 ms·

In 2015 I quit my full-time job as an ASP.NET developer to build what I think of as an "atlas of history" for the Italian Renaissance. It is a semantic-web styl

In 2015 I quit my full-time job as an ASP.NET developer to build what I think of as an "atlas of history" for the Italian Renaissance. It is a semantic-web style database build in Neo4j, .NET MVC, and KnockoutJS, and is an attempt to build a map of historical events and personalities for the digital humanities. http://the-codex.net http://the-codex.net I am currently the sole developer, product designer, and researcher on the project -- but I am looking for collaborators who would be willing to help me take this further. As an "atlas of history" is a broad concept I decided to give the project clarity by focusing primary source documents from the Italian Renaissance. The two main sources at present are the 'Florentine Diary' of Luca Landucci and the letters of Michelangelo. I have entered about 40 years worth of entries from Landucci's diary and a good portion of Michelangelo's early letters from his Roman period. In the process I have added hundreds of historical personalities, places, artworks, etc., in order to give the user real data to work with. I have also built various screens with data visualisation tools to mine the historical events. And of course I have built an extensive back-end for managing the data and relationships. Is anyone interested in helping me out? I'd love input from anyone with an interest in art history and graphic design, or data visualisation tool, Neo4j, or anyone who wants to help me research and enter data. Feel free to email me any time at: iian.d.neill@gmail.com In the meantime, why not check out the Control Panel on Leonardo da Vinci's dataset. Clicking any links in the text will load the datasets for those entities; or you can search for them by name. Why not try adding Michelangelo's dataset to the mix? You can then switch between the three data-vis modes at the bottom, fiddle with the date filters, etc. http://the-codex.net/Time/ControlPanel http://the-codex.net/Time/ControlPanel Many thanks, Iian

goldfeld 11y ago
I appreciate what you've built and I'm interested in helping out. What is the way going forward for the project? Do you plan on monetizing it? Is code eventually gonna be public?
- argimenes 11y ago
  Hi, Thanks for your interest! I have no plans for monetizing the project, partly because I cannot see a meaningful way to do that, but mostly because it is a research endeavour. The code is currently public on my BitBucket repo, and I'll shoot through the URL as soon as I can. Basically, I want to explore the limits of how graph databases can be applied to the visualisation of history. My immediate goals are to: (A) expand the range of datasets that are collected; (B) improve the tools for entering and annotating the data to make the process quicker and more automated; (C) to add more powerful data visualisations; and (D) to complete the input of the primary source datasets mentioned and add more. I want to build a graph database time machine of the Italian Renaissance so you can pick a day in history and see what was happening across Italy with various sources, see events plotted or even animated on a map (e.g., watch a battle unfold or even weather events), etc. The system is underpinned by subject tags which are taxonomically organised in a 'is-a' hierarchy, which I hope will provide a 'semantic zoom in/out' functionality. I understand that the current platform of C#, ASP.NET MVC, KnockoutJS may be a turn off ... And I am open to other technologies if that is crucial. But conversion to another platform would certainly take a lot of time and manpower. But to be honest I think the hard work is mainly in the product design, the visualisations, the data-entry, and the Neo4j Cypher queries. What kind of collaboration did you have in mind?
  - goldfeld 11y ago
    I could whip up an alternative frontend implementation if you'd like, I've been looking for a project where to flex my ClojureScript/Om Next (based on React) muscles. I'm very experienced both with Clojure and with the frontend, and with interface design. I actually worked with KnockoutJS (in a C# shop I used to work at) many years ago, but yeah I don't have fond memories of MVC/MVVM for large projects/efforts, the React model speaks more to me.
    - argimenes 11y ago
      ClojureScript and React sound interesting to me, too, and might be a good fit for the rich UI Control Panel page, which will need to handle large JS data-sets and update visualisations quickly. I've been finding that KnockoutJS is a bit sluggish when wiring up the filters as computed function observables. It could be something in my filter code, also, but I have seen benchmarks where KnockoutJS doesn't perform well in. We should talk further ... drop me a line at my email? iian.d.neill@gmail.com

goldfeld 11y ago

I appreciate what you've built and I'm interested in helping out. What is the way going forward for the project? Do you plan on monetizing it? Is code eventually gonna be public?

argimenes 11y ago
Hi, Thanks for your interest! I have no plans for monetizing the project, partly because I cannot see a meaningful way to do that, but mostly because it is a research endeavour. The code is currently public on my BitBucket repo, and I'll shoot through the URL as soon as I can. Basically, I want to explore the limits of how graph databases can be applied to the visualisation of history. My immediate goals are to: (A) expand the range of datasets that are collected; (B) improve the tools for entering and annotating the data to make the process quicker and more automated; (C) to add more powerful data visualisations; and (D) to complete the input of the primary source datasets mentioned and add more. I want to build a graph database time machine of the Italian Renaissance so you can pick a day in history and see what was happening across Italy with various sources, see events plotted or even animated on a map (e.g., watch a battle unfold or even weather events), etc. The system is underpinned by subject tags which are taxonomically organised in a 'is-a' hierarchy, which I hope will provide a 'semantic zoom in/out' functionality. I understand that the current platform of C#, ASP.NET MVC, KnockoutJS may be a turn off ... And I am open to other technologies if that is crucial. But conversion to another platform would certainly take a lot of time and manpower. But to be honest I think the hard work is mainly in the product design, the visualisations, the data-entry, and the Neo4j Cypher queries. What kind of collaboration did you have in mind?
- goldfeld 11y ago
  I could whip up an alternative frontend implementation if you'd like, I've been looking for a project where to flex my ClojureScript/Om Next (based on React) muscles. I'm very experienced both with Clojure and with the frontend, and with interface design. I actually worked with KnockoutJS (in a C# shop I used to work at) many years ago, but yeah I don't have fond memories of MVC/MVVM for large projects/efforts, the React model speaks more to me.
  - argimenes 11y ago
    ClojureScript and React sound interesting to me, too, and might be a good fit for the rich UI Control Panel page, which will need to handle large JS data-sets and update visualisations quickly. I've been finding that KnockoutJS is a bit sluggish when wiring up the filters as computed function observables. It could be something in my filter code, also, but I have seen benchmarks where KnockoutJS doesn't perform well in. We should talk further ... drop me a line at my email? iian.d.neill@gmail.com