Archive for the ‘Technology’ Category

Google Reader: Workaround for sharing

As we are all aware Google Reader disabled sharing other than on Google+ in it’s new version. This obviously upset me immensely because I used my share feed for my blog. After trying out numerous suggestions, this is what worked:

1. Subscribe to your original shared feed
2. Pray to God you had created a public tag prior to the switch-over to the new reader and despite whatever you named it, use it! Newly created tags are by default private and there is no means of making them public at the moment.
3. Using the old tag, create a bundle. There should be drop-down to the right of the tag name in the left pane.
4. Also add/drag the subscribed shared feed into the bundle.
5. Change all your blog widgets using the original Shared feed RSS to the RSS feed for this new bundle.

I tried using Chrome extensions to display the original shared icon but that has a number of issues:
- The extensions stopped working for me soon after
- My email got hacked into around the time I got the extensions. It could have been because of insecure scripts
- Google is going to eventually remove the original sharing from its API so might as well stick with something they’re not going to scrap soon. Or one would hope so.

Hope this helps some fellow frustrated soul.

  • Share/Bookmark

Converting ODK XML Data to CSV

Context: Household survey in rural Tanzania

Task: Combine data from XML files and display in an Excel File

Challenges:

  • Nested loop within survey which required questions to be repeated so survey and questions were not one-to-one
  • Number of fields for survey exceeded 255
  • Each survey was in a separate file
  • >350 submission files so need mass combination and conversion

Data Provided:

  • Survey in XML format for ODK Collect
  • Submissions in XML format using ODK Collect. Naming of submission files based on survey name, data and time collected.

The following methods were attempted in extracting data. Some successful, some not, with clear reasons as to why not.

Method 1: Use Excel

Excel’s XML to CSV or XLS function is very easy to use but issues were:

  • Each XML file is saved to a separate CSV file so does not aggregate data.
  • Does not take into account nested XML and will only add in columns for the primary level in XML

Method 2: Use MS Access

The idea was to first import into Access and then to CSV. Importing to Access worked exceptionally well. It automatically creates related tables for nested XML tags. However issues were:

  • The size of the survey itself was too big to store in an Access table. It would have taken far too long to change data types one by one. In addition if the table was broken up, the automatic feature for import would not work. There is probably a way to do all this in Visual Basic but again time constraints.

Method 3: Install ODK Aggregate and use inbuilt stable version of ODK Briefcase:

ODK Briefcase is the companion tool for ODK Aggregate to import and export data from ODK generated files. However the inbuilt version could only:

  • Import one file at a time
  • ODK Aggregate is a little complicated to install. You can either install it on Google App Engine or locally but the installation procedure is a tad confusing.

Method 4: KoBo Post Processor

KoBo Post Processor is a third party tool to aggregate data from XML files. However the issues with this were:

  • Could not read the nested XML data
  • Had it’s own algorithm for displaying columns and did not follow the survey format

Method 5: Stand Alone Beta Version of ODK Briefcase

Using ODK Briefcase v1.0 Beta 2.jar is the method which worked but since there wasn’t much documentation on this, it took awhile to figure out and find. There are only a couple of threads on the ODK Developers Google Group which mention this. There were obviously a ton of issues considering it’s in Beta mode but after a number of tweaks and re-positioning data it worked. Issues faced during this process and remedies:

  • ODK Briefcase requires it reads the data from a very specific directory structure. It needs to be the exact same as that created by ODK Aggregate for forms. Example of the directory structure in the ODK Briefcase folder:

    Forms

    Tanzanian Survey

    Instances

    TanzanianSurvey-11032011-1203

    submission.xml

  • It is important that the XML data file be named submission.xml which is only possible if each submission is in a different folder. Unfortunately our data was not named submission.xml so I used Bulk Rename Utility. There are other ways to bulk rename in Windows but the search and replace using F2 method adds a number in parenthesis which would not work.
  • Regarding the nested XML, since there were only 37 files where this was the case, we went through each file and flattened it to one level. However Windows doesn’t let you search within XML files so you’d probably need to download some software like XML Search

Ideally all of this could have been done by writing a simple script and writing to the same CSV file. However I was aiming to find something which could have been replicable by a non-tech individual. Looking forward to the next version of ODK Briefcase.

  • Share/Bookmark

Net Change 2010: Mobile Tech

Wednesday, 9th June was spent attending the Mobile Tech portion of the Net Change 2010 conference here in Toronto. Following are some highlights and notes related to it:

Mobile Takeover: How cell phones are leading to economic and democratic progress
Professor Iqbal Z. Quadir (Founder, GrameenPhone)

  • Professor Quadir works at the Legatum Centre for MIT which encourages entrepreneurship in developing countries as a means of economic growth.
  • “For-profit is a healthy thing”
  • “Concentration of resources is related to economic degradation” For example concentration of power.
  • “Governments managing people” versus “People managing governments”. Democratization usually leads from the former to the latter.
  • England’s move towards a more egalitarian state started from tax collection
  • “The history of liberty is a history of the limitation of governmental power, not the increase of it.” – Woodrow Wilson
  • Commented on the title for The Elusive Quest for Growth: Economists’ Adventures and Misadventures in the Tropics by William Easterly. He was referring to the fact that economists can get away with being deemed adventurous where as other professions, such as say doctors, couldn’t possibly refer to their forays in health as adventures. Showing a stark difference in how we perceive health and poverty. Not much of a feedback loop when it comes to economic endeavors in the developing regions.
  • “Connectivity is productivity”
  • Myth debunking: “Poor people are resources” NOT “Poor people are recipients”
  • Recommended reading: “Global Heroes”, The Economist; “The world turned upside down”, The Economist; “The Other Greeks: The Family Farm and the Agrarian Roots of Western Civilization” by Victor Davis Hanson

Katrin Verclas, MobileActive.org: Mobiles and Non-Profits

Lisa Campbell Salazar, TakingITMobile

This was a remote presentation via Skype. Was a little difficult to follow but I got a better idea of what the aim was when I visited the project website. Mostly gathering data from a number of developing countries regarding how youth use mobile technology.

Michelle Hamilton Page Toronto Public Health, Proximity Marketing

  • Hyper Tag: A wearable device which can broadcast / transfer information via blue-tooth. Michelle spoke about past and future campaigns utilizing this device. The phones receiving the information have to be set to discoverable and the user has the capability to accept or reject a transmission request.
  • inspot.org: STD notification / encouragement for testing e-cards. This was a service one of the Toronto Public Health campaigns was promoting. The conversion rate (in this case the number of phones observed versus phones reached) was fairly high. However there seemed to be more of a focus on dissemination and less on the final goal. Michelle mentioned though that they are planning to improve their analysis methods.
  • UrbanDNA: The company Toronto Public Health chose to partner with for this particular kind of campaign.
  • Observation: Mass distribution did not work well for them and they found one-on-one got a lot more people involved.
  • Short Code woes: Need to make sure you own them. Could end up in a situation where you lose them after they’ve been established as branding or marketing contact points.

Kelisha Peart, Black CAP

  • Case Study: Focus groups in targeted regions within the Greater Toronto Area
  • Interesting results. See website for more details

Hima Bhatavia for Jim Monis, Mobile Giving Foundation:

Company started in the United States and from what I gathered they proceeded to form partnerships with mobile carriers which allow users to use short codes to donate money to charities. The donation shows up on their cellphone bill at the end of the month. The Mobile Giving Foundation charges each charity around $350 to sign up and sends out checks every 60-90 days. Currently there seem to be a number of partner organizations which facilitate the process such as ZipGive.

Andrew D’Amico, ZipGive

  • Currently running the Young Artists for Haiti campaign
  • Maximum donation cap: Set to $30 since if a user chooses not to pay their bill, the carrier gets saddled with the cost.
  • However there is the potential to activate monthly donations or a multi-layered pledge option for larger donations.
  • Does not have GEO tagging capability (Important consideration pointed out by Dale Zak)
  • Currently all text outs are handled by ZipGive. They have their own databases of users which can be populated by data you provide. I believe the software has the capability to allow companies to send out their own text outs, but for the moment it’s centralized to prevent abuse. But I think down the road there should be more control within the individual companies. It would probably cost less and save time spent communicating with the middle man.
  • Feedback loops need strengthening i.e. a means of co-relating texts sent out with number of users who followed up on them
  • Premium text messages: One of the audience members wondered why those are not being used instead of the short codes. Apparently there isn’t a cap on the donation amount in those cases. According to the presenters, the charities do not get a 100% of the proceeds of premium text messages and some carriers have chosen to disable this option since it has been known for abuse.
  • SMS vs QR Codes: Unfortunately I don’t know much about QR codes but according to the presenters these tend to be restricted to applications within smart phones. If any of you have any addition information, feel free to use the comments sections of this post.

Measuring Success on Mobile, Katrin Verclas (mobilebenchmarks.org)

Kartrin recently launched a website which is related to collecting, analyzing and disseminating information about metrics related to using mobile technology for advocacy, fund raising etc. MobileActive partnered with M+R Strategic Services for this project. Some interesting things which came up during this conversation:

  • Keyword opt-in is really low
  • Even though you may have blocked your cellphone number so that people can’t see it when you call, it will still be displayed in you text message them.
  • Important to consider churn rate (rate at which mobile numbers go bad)
  • A re-opt in is required if you switch carriers even if there is number portability
  • Incentive to participate can be good e.g. free airtime.
  • Emphasis on the FailFast concept i.e. by all means try but know when to quit.

Mobile Volunteerism Dale Zak Apps4Good.ca

I could not make it for this one since it was at the same time as Katrin’s but be sure to check out the website and I’ll try and post up any related presentation material as soon as I can get hold of it.

Workshop: Building Mobile into My Communication Strategy, HJC

By this point I was a little wiped out. There were some interesting projects people were working on but all in all I don’t think many people had come with a specific project in mind. Most seemed to be considering the potential.

Random Observations at the Conference:

  • Everyone had a Mac with the exception of one Thinkpad and my Dell.
  • Majority had iPhones and Blackberrys. Katrin and I seemed to be the only ones with Android based smart phones.
  • The digital installation was AWESOME. It was called SEED and the idea was that you could call in a local number and a seed would appear on the screen which was attributed to you. You could then use phone keys to grow and spread your tree. After you were finished, your virtual tree would be added to the virtual forest and Evergreen would plant a real tree in lieu of the one you created virtually.
  • Marigolds & Onions: I really liked lunch so here’s a shout out to the caterers.
  • Twitter tag for the conference: #ncwk
  • Last but not least shout out to the organizers for a great event!
  • And yes I know there are not enough pictures primarily because I forgot to take along my SLR.
  • Share/Bookmark