Ideas and things I have been working on in no particular order.
An initative to crowd source an open data geolocated version of the Swedish post code system. We developed a website, apps for iOS and Android and produced datasets. Over 1000 people were involved in collecting addresses and post codes. We delivered the results and suggestions to the Swedish Post inquiry (SOU 2016:27)
Tableau – Kolada Web Data Connector
Using KPI data about Swedish municipalities in Tableau can be difficult. This is a web data connector that makes it easy to load Kolada data in Tableau.
Pimmer – image mining PDF document collections
Sometimes you need to extract non-text content from collections of PDF files (e.g. all hat images from a book about hats). Pimmer analyses PDF documents for image features using Canny edge detector in OpenCV and exports a collection of image files. Packaged as a docker application.
Flash Card Ontology
The Flash Card Ontology is a specification for flash cards which is intended to be used as an aid in memorization of information.
Munin - a social media archiver
Munin is a tool to archive open Facebook, Instagram and VKontakte account posts. Data is archived in the WARC file format. A playback tool and a simple dashboard is available to monitor collections.
Swedish political advertising on Facebook
Political advertising on Facebook is at risk of being lost forever. With the 2018 election in Sweden I was worried that there will be little to no trace left of campaigns and sponsored political content after the election takes place. I translated the ProPublica Facebook ad collector to Swedish and have set up an archive of Swedish political ads. Site had to move offline after copyright claims from a newspaper about some of the images political parties were using in their ads (and thus reflected on my collection website).
textamig.se - shorten time for captioning Youtube and Vimeo videos
Textamig.se (in Swedish) is a simple tool to help people get started with captioning their Swedish online videos. Captioning improves accessibility and usability but it can be difficult to get started. Youtube has automatic captioning for English but many other languages have to manually provide their own captioning. This tool uses Google speech recognition (via autosub) to get a basic time coded subtitle file.
Userpoll.io - empower users to evaluate software
Userpoll.io enables anyone run a System Usability Score (SUS) survey. Many organizations lack in-house user experience specialists and have no resources to hire consultants. Conducting a usability survey with a limited number of users can help you motivate why further research is necessary. The goal is to help non-UX people to take the first step in making software better.
Spreadskos/skos2html - a spreadsheet to SKOS converter
A simple Ruby gem to enable editing of SKOS vocabularies in Excel and other spreadsheet software. Check out skos2html for a gem that converts SKOS to a human readable HTML file.
Eurlex.nu - a semantic web version of European legal information
eurlex.nu (now offline) was an experiment to see if european legal information can be put online using semantic web techology. Eurlex.nu held some 20,000 documents in two languages or more. Documents were presented as XHTML 1+RDFa and there was a SPARQL interface for queries.
opengov.se - open government data access in Sweden
Opengov.se was an private initiative to highlight available public datasets in Sweden. It was the first open data portal in Sweden. It contained a catalog of government datasets, their formats and usage restrictions.
The goal was to highlight the benefits of open access to government data and explain how this is done in practice until the official government portal came online.
Eutveckling.se - guidelines for and evaluation of public sector websites
An online version of the Swedish National Guidelines for Public Sector Websites. I co-authored the guidelines from 2004 to 2008 and infrequently in the years after. The guidelines were developed by Verva, the Swedish Administrative Development Agency and are today hosted by the Swedish agency for digital government, DIGG.
Twoot - an OS X twitter client based on JQuery and Fluid
An exploration of a site specific browser (SSB), using Fluid, in order to learn more about jQuery, json and git.
Fangs - the screen reader emulator plugin for Firefox
Firefox extension that renders text representations of web pages identical (almost) to one of the more popular screen readers. The idea is to help developers get a better understanding of how an assistive technology interprets their markup. See the Fangs homepage for more details. Had thousands of users but no longer works with the latest versions of Firefox. Fangs project page and help file at standards-schmandards.com.
Ruby Accessibility Analysis Kit
RAAKT was a web accessibility analysis library written in Ruby. The idea is to help web developers catch basic accessibility issues as part of the development process instead of after-the-fact testing. RAAKT integrates nicely in automated testing frameworks. For more information see the Raakt wiki.
It was used in an online version of RAAKT by the company Netrelations and by several government organizations.
SimpleCrawler was a basic web crawling library for Ruby. Single threaded and intended for basic crawling of websites.
- Source: http://simplecrawler.rubyforge.org/svn/
- Documentation: See the Simplecrawler wiki.