A two-day analysis of #stayhomesavelives

Covid-19 has disrupted everybody’s lives. Parents working from home (when lucky) or waiting to go back to work (when unlucky). Children are being home schooled and schools are silently trying to cope. It is tough. What is tougher is the fact that the UK has officially surpassed Italy and is now trying to show that they are no so worse off than Spain. Needless to say, the UK had more time to prepare for Covid but this is not the point of this post.

People have shown resilience, kindness and hope. Local networks of volunteers and businesses have shed a new lights on how people can establish communities of support in hard times, even when institutions cannot support. I will write another post about some good examples in Central Scotland. Most importantly, the NHS staff demonstrated that decade-long cuts have been unable to erode their dedication to their patients. And people have appreciated this. Twitter has been inundated by messages of support to the  NHS and the use of the #COVID-19 as well #stayhomesavelives have abounded.

The latter one has been analysed on twitter and two-day worth of tweets have been studied looking at the key tags around which a network of words, people and thoughts have gathered.

The tweets of two days gathered, created two main clusters: keeping safe but also going through hardship and patiently await the end of these hard and difficult times. What is interesting is how the network of most used words (or toptags) revolve around two major categories: the togetherness (which  is also including the NHS) and the self-isolation aspect of such togetherness.

The discussions (and these are only two days of tweets!) are very polarised toward everyone being in this together and the necessity to be thoughtful and the flip side of keeping millions isolated at home. Key terms such as hardship and patience emerge just as insistently as self-isolation (right hand side of the network), family and fun. The network (terrible rendition on the blog post unfortunately) is split in two clusters.This network could potentially help us understand the complexity of these tough times and the lessons we can learn from these discussions. For decades we have been talking about the positive aspects of technology, how it saves us time, how we can be in a videocall and still wear out pj bottoms while our line manager is talking about very important strategies. What nobody could ever foresee if how the physical distancing that a catastrophe like COVID-19 caused, could bring us so socially close, appreciating the multiple communities we take for granted. The NHS is the only such community that emerges in these 49,000 tweets in the two crucial days during which  the UK reached, and sadly surpassed,  their 30,000th death.

Toptag network -#stayhomesavelive

One cluster, the one on the left, is about family and being static, staying at home. The opposite cluster, the one on the right, is about self-isolation, wait and toughness of these unprecedent times. People, care and NHS stay right at the heart of this network of tweets, indicating (perhaps corroborating) how this lockdown has been hard on physical distancing but it has certainly brought (back) to light the important role of social closeness, belief in certain institutions like the NHS and the sense of togetherness.

What does this mean for non profit organisations, institutions but also governments?

The networks of support that have emerged during this pandemic have been phenomenal. Facebook pages mushroomed (too bad Facebook didn’t let me mine these beautiful pages), messages of support and networks of volunteers ready to give up their time, offer to do grocery shopping for the elderly and leave the shopping at the doorstep (so no thank yous involved), people that helped other people paying bills, sharing grocery shopping budgets, calling on one another for help without fears.

This network rendition of the discussions taking place on Twitter using #stayhomesavelives tell a story where resilience mixed and enmeshes with fear and stories of people helping -and chatting with- people. Plain and simple. These are hard times and there is no trivialising them with a network of tweets. Tweets obviously do not tell the stories of casual or seasonal workers, small businesses and bnb that have received en-masse cancellations. However, these solidarity networks that have emerged and have found space on social media should be carefully studied and preciously kept because they show that communities-driven solidarity projects, micro-credit and small incentives through communities (in this case virtual) of people that have finally realised to be on the same boat are more powerful than any top-down initiatives aimed at supporting some of us in these difficult and trying times.

Let’s talk about #DownSyndrome

I have been working on a research project. The PI approached me because she needed someone that could do some digital story-telling. So, here I am trying to figure out how to put a rather specific research question into a compelling communication project. Something people would want to know about.

The project is about dementia in people with learning and cognitive disabilities, specifically Down Syndrome. A massive topic I would say. Dementia: hot topic; Down syndrome: complex topic.

A first step to understand how to communicate is to understand what is that will be communicated. So I started researching dementia AND, separately, Down Syndrome. I wanted to understand the sentiment, the tone, the feelings towards each of them. What campaigns have been run to raise awareness on dementia? Are there campaigns to support or raise awarenes on Down Syndrome? What are the tones?

Firstly I approached colleagues in the Film and Media Section of my department to know if there is any literature on the stereotyping of the person with Down Syndrome and if they had ever read anything about a sitcom (cannot recall the name now!). Then I thought of turning to what I m most interested in: social media. Twitter is the fastest and easiest avenue of research and I undertook some sentiment analysis of the campaign #DownSyndrome. I randomly selected over 2000 tweets to perform an initial sentiment and frequency analysis. And here came the surprise.

What have I noticed? 70% of the tweets were very positive. Overwhelming correct? Sure. Then I went into the remaining 30ish %. There emerged a little sarcasm (see the picture).Picture1


The sarcasm couldn’t really be analysed with the code so I have looked a bit closer to the tweets I had downloaded. There emerged a completely different matter. The sarcasm was not directed toward the campaign or the person with Down Syndrome; not toward other people that had used the # or the issue altogether. A quarrel between those that had associated the #DownSyndrome to the #proLife and those that had, instead, associated the #DownSyndrome to the #proChoice. The tweet that unleashed the angry tweets was a link to this video.

The discussions therefore moved from the person with Down Syndrome (actor) to a much more complex issue: the right to choose whether or not a person with Down Syndrome shall come to this world; the right of the mother, the right of the child. And the tweets were rather hard core, accusing one another…losing completely sight of the issue at stake: the awareness campaign.

Of course, the positive tweets also offer very interesting insights; following upon Van Gameren-Oosterom research on the stereotyping of the person with Down Syndrome, the positive tweets certainly meant good but tended to portray a positive, smiling child, when the reality might be a bit different.

Therefore, a series of questions arise. Are we stereotyping a segment of the population that has never really been the centre of much media research? Are we using one issue (in this case an awareness campaign) to bring forward other battles (i.e. prochoice or prolife discourses)? If so, what are the scenarios that we must consider when communicating to the world that given the life expectancy of people with Down Syndrome has now reached (and exceeds at times) the 70 years of age and this under-researched segment of the population will need to be made aware of the issue of dementia?


CJ “RE-Classification” Project

In 2006 geographer Keith Yearman launched the “Juarez Declassification” Project aimed at showcasing the attempts to make the issues of feminicidios disappear from national and international agendas. The project contained cables from US and Canadian Embassies with regards to the “bodies found dumped” in deserted areas in the city. The project was dismissed and abandoned.  There have been other attempts to merge computation, statistical evidence and the issue of feminicidios in CJ; an example could be the project “Lucy Project” a Phd project that tried to gather information on the missing girls in Ciudad Juarez and offer a platform that could be used to report and denounce women disappearances and murders in the city. Machine learning and computation met again with Ciudad Juarez with various attempts to link the presence of gang graffiti in the city and the drug routes. All of these have been either abandoned or have lacked in-depth understanding of the city and its inhabitants.

The idea of the “CJ RE-classification project” is not to reinvent the wheel but to archive what has been done, over the years, to denounce, report and put an end to the forced disappearances and violent murders of women -and girls as young as six years old- in the city. The aim is to gather evidence: if it is not possible to know how many, at least it is possible to know how much effort has been put into the making and breaking of feminicidios. If in fact on the one side mothers march to know who is taking responsibility, on the other academics and researchers have accused feminicidios to be “myth”.

Rather than dismiss the various positions and privilege one view or truth over another, the idea of the project is to put everything in one place an understand the narrative, the breadth and depth of the issue and the ways in which the issue has been narrated (either to make it a “myth” or a major problem the city has been dealing with for over two decades).

After two years of building the database and structuring the information scattered online with regards to feminicidios in Ciudad Juarez, it has finally been possible to create a Google form directly linked to a SQL database (no chance to access the DB from the form though so info is safely stored). The form allows to upload documents such as pictures, videos and even audio files (for podcasts that are not already shared online). These will be identified with a unique ID and will enrich the already rich database. The layout is very basic but it does the job.

Screenshot 2017-08-30 08.55.27

What is the purpose of the database? It was born as a way to record and analyse the image narrative around feminicidios in CJ. Little by little it turned into a necessary archive of information and multi-media material that could record, in one place, the scattered information on feminicidios in Ciudad Juarez. Thanks to the contribution of various activists and friends, it is growing and it has come to include academic papers, conference proceedings, university projects (such as google maps that visualise the number of bodies found in specific locations or years), podcasts, and much more.

The internet offers a rather rich repository of images and audio visual material, “clues” (Eco, 1984) that can inform on the efforts that activists, feminists and families of murdered and missing women in Ciudad Juarez have put in making the issue of female homicides and disappearances visible.

Why a database if the Internet already offers a lot of information? Because no machine learning algorithm will be able to cluster and make sense of the complex narrative associated with the issue and offer some statistical significance to a problem that has never been quantified and has always remained a very local and localised tragedy.

Anybody can submit the form and contribute to the database, which already comprises of over 5,000 between academic papers, pictures of protests, newspapers articles (mainly online newspapers), videos and a couple of podcasts. But we can always do more!

Screenshot 2017-08-30 09.03.24
Image correlation analysis


To contribute information and keep on enriching the database, please fill in the FORM.

Database-ness and visual narratives

The UK tradition of Software studies seems to be a far away theoretical world, where Deleuze merges with SQL and Basics programming. I have used it as a point of reference to understand how the activities of “bordering” (Lury, 2014) can be applied to the cases of missing, disappeared and murdered women of Ciudad Juarez. The hope is to map the bordering activities of those ‘underdog’ that seem to have no statistical relevance for the system, any system.

Bing, Google and Yahoo! have been investigated following the logic of keywords (there seems to be no other way of researching the wealth of information on these beautifully tailored search engines) using two geographical denominations: .com & .mx

The visual narrative has been built following something that is intriguing in visual computing or whatever it is called: object saliency. For now no algorithm has been used but our eyes have been used to detect: the salient object, the context of the image and, of course, the image “container”-> platforms.

Platform and researched keyword have been inputted and the results are now going to be analysed with the hope to automate the whole thing at some point. I m tired of saying very soon cause nobody ‘volunteers’ time and I had to learn R, MySQL, Gimp and what not only because nobody shares the sense of duty that academics should generally have!

The database now looks like this Fig.1 and the results, which are all plotted in R for the time being and look horrible because I m still learning multi-axis plotting and playing around with eucledian distance (so terribly western and masculine!)/k-means and what not are Fig.2.

The paper will follow. The results have also found out 10,000 links on the TOR browser that show the ‘fun’ of killing women in Juarez. All have been reported of course!




Feminicidios en Ciudad Juárez


Ciudad Juárez es la ciudad de México en la que puede decirse que el crimen organizado y la delincuencia campan a sus anchas, mientras que el Gobierno y sus instituciones perecen asfixiadas entre la brutalidad y la barbarie. Sus ciudadanos se enfrentan a la corrupción, el encubrimiento y la indiferencia de los funcionarios.

La ONG Nuestras Hijas de Regreso a Casa trabaja hace años para aclarar la muerte y la desaparición de centenares de mujeres, que creen poco creíble la explicación que da la fiscalía sobre el tráfico de personas.

Hace ya muchos años que las mujeres en Ciudad Juárez  no pueden vivir tranquilas, muchas de ellas desaparecen  y nunca más se vuelve a saber de ellas, aunque a veces las encuentran muertas después de haber sido violadas y maltratadas brutalmente, con partes de su cuerpo mutiladas o quemadas.

El Gobierno no se preocupa por tales hechos ni nadie se…

View original post 260 more words

A must read: Ecologies of Power

Just caught up with Ecologies of Power by Pierre Bélanger and Alexander Arroyo (MIT Press), which – as the subtitle reveals – is a fascinating countermapping of the Pentagon’s logistical landscapes and military geographies: This book is not about war, nor is it a history of war. Avoiding the shock and awe of wartime images, […]

via Counter-mapping and ecologies of military power — geographical imaginations

From the Guardian (20.11.2016) “Boko Haram’s forgotten victims return to a humanitarian disaster”

Counting of the missing(s)


And we thought what is most scary of the whole article is the following:

“Unlike the refugees from Syria and elsewhere, whose perilous journeys to Europe have rarely been out of the headlines, the people of north-east Nigeria have remained almost entirely below the media radar.

Even in Nigeria, there has been little coverage of their plight and with the country’s economy now in recession for the first time in 25 years, mainly because of the collapse in oil prices, the government has been unable to allocate enough resources to provide the food and other aid that they need.”


Worth reading the whole article.

They have a name: the cartography of feminicidios in Ciudad Juarez

Yesterday a friend sent me a link to an article about feminicidios. An interactive map that had patiently been created aggregating data from several sources:the archives of the

Fiscalía Especializada en la Investigación de Asesinatos de Mujeres de la Subprocuraduría de Justicia Zona Norte, the Comisión Nacional de los Derechos Humanos y de organizaciones no gubernamentales.

The team is made by  Diana Washington, Sergio González Rodríguez, Rohry Benítez, Adriana Candia, Patricia Cabrera y Julia Monárrez, and the data gathering has used news from local newspapers, non governmental organisations and, overall, some crucial blogs. The victims have been given a name; their stories have been told and they have finally emerged.

despite a general sense of happiness, there also emerges a profound disappointment at the lack of interest that one of our project, namely giving the names to such victims, has met these years.

However, we don t want to spoil such a good beginning (hoping this project will expand and perhaps collaboration emerge). The link is below.



POV in the killbox — geographical imaginations

An update on Joe DeLappe‘s Killbox project (my original post, with links to more info on the concept of a killbox, is here). Over at Quartz, Ananya Bhattacharya provides more details about the latest iteration of the simulation: Killbox, an online two-player game named after the military term for an area targeted for destruction, serves as […]

via POV in the killbox — geographical imaginations

Emerging connections: the affinity map

Last night the ENAC (broadly speaking, the Faculty  of Architecture at the Swiss Institute of Technology in Lausanne) had the pleasure to reflect on their own collaborative networks. Through the realisation of a bespoke project named “affinity map” realised by my friend and colleague Dario Rodighiero , the ENAC had a first touch-and-feel of how their policies of collaboration and mutual support have (or have not!) worked.

Dario’s main objective is to “make the invisible visible”. In times where data and big data pressure businesses and academic alike, Dario’s tool is inspiring; it has certainly forced each component of the map to reflect on their own work and behaviour. We can picture some relieved faces of lecturers and researchers (“yes! I have loads of connections!”) and some others a bit more ashamed. The affinity map is based on the process of gathering and meticulous organisation and display (visualisation) of ‘messy data’, coming from people’s work. The map is much closer to reality than the shiny world of Data Science wants us to believe for we deal quotidie with messy data.



The affinity map showcased last night makes us rethink:

  1. Governance
  2. Performativity


One way or another, we tend to be under pressure; we need to collaborate. Before, colaboration would very much be interpreted as weekly meetings with the team at work, monthly meetings with the division and perhaps Christmas parties with the company/institution. Now, collaboration is becoming part of KPIs and most certainly, there is no project and no institution that doesn’t strive for the best possible results through collaboration. We are being asked to collaborate to win prizes, collaboratively review RFPs, submit inter-disciplinary and collaborative research proposals to increase our chances to win etc… Yes, we are under  pressure, and changing the habit of ‘pretending to collaborate’ to a habit of ‘true collaboration’ is tough.  There is very little knowledge on how to monitor collaboration. Dario’s visualisation, hopefully, will become a useful tool to understand the full extent of collaboration and, also, understand how it also translates into efficiency and competitive advantage. It might sound scary but, willingly or not, we have embraced business logics of efficiency, exposure and global reach.



How can data perform a new reality? This question is far from clear. However, what we see in this very quick and simple video isn’t only the translation of some very complex data into a nice dataviz project; it is also a whole new way of understanding the working of a small world, a lab, an institution or a company. The lack of collaboration or the wealth of networks of collaborations are translated into an algorithm that codifies the institutions’  ‘willingness’ to collaborate. The tool is performed through data; the academics’ were forced to look reflexively into their own practices. The affinity map certainly raises tough questions and it certainly brings up possible discussions on the discursive formation of collaboration and can have unknown -disciplinary- consequences (more policies?).

Networks of academics, their works and their networks of supervisions is a fascinating matter that opens the door to reflections over the impact of such a tool on universities’ research and teaching frameworks and policies, on governance and, most certainly, on the ethics of transposing collaboration and inter-disciplinarity into an eye-catching visualisation. All in all, a pretty awesome job



” Lack of data on sexual rights leaves millions of girls “invisible” – report”

********Article copied from Reuters Foundation*******

Author:  Lin Taylor

We are reposting an article on lack of data on sexual rights as published on the Reuters Foundation Website on Monday, 3 October 2016 00:00 GMT.

“LONDON, Oct 3 (Thomson Reuters Foundation) – Millions of girls are left “invisible” because of a lack of data, a children’s charity said on Monday, and the absence of accurate statistics on issues such as sexual violence means policymakers cannot draw up effective plans to help them.

There is no data that fully captures the daily realities of girls in poor communities, Plan International said in a report, including why girls drop out of school or how many become pregnant because of sexual violence.

“We do count how many girls start school, but we actually don’t count how many girls leave school,” said Anne-Birgitte Albrectsen, CEO of Plan International.

“We also don’t have any data on why they leave school – whether they were forced into marriage, whether they became pregnant, or were sexually assaulted at school,” she said in an interview with the Thomson Reuters Foundation.

“It’s leaving a lot of girls invisible.”

Globally it is estimated over 2 million girls under the age of 15 become mothers each year, but the number is uncertain as official data tends to only track births of women aged 15 to 49 even though girls can get pregnant from age 11 or so onwards.

UNICEF estimates around 150 million girls around the world have been sexually assaulted.

Since talking about sexual violence or reproductive rights remains a taboo in some communities, Albrectsen said, collecting accurate statistics on these issues is the most challenging, as there are “political and cultural hurdles to overcome.”

Without data, intervention programmes cannot effectively improve the lives of girls who are most at risk, she added.

The report comes a year after world leaders agreed on an ambitious new set of global goals designed to improve lives in all countries by 2030.

The U.N.’s 17 Sustainable Development Goals with 169 targets are a roadmap to end poverty and hunger, fight gender inequality and conquer climate change over the next 15 years.

Albrectsen said governments must invest in data collection, and capture meaningful statistics that reflect what girls in their communities are facing, such as pregnancies, rape and school drop-outs.

“With the availability of data, the lived realities of girls will become a lot more visible to policymakers,” she said.”

(Reporting by Lin Taylor @linnytayls, Editing by Ros Russell; Please credit the Thomson Reuters Foundation, the charitable arm of Thomson Reuters that covers humanitarian issues, conflicts, global land and property rights, modern slavery and human trafficking, women’s rights, and climate change. Visit http://news.trust.org to see more stories)

A WOMAN NAMED ELLEN — The Neighborhood


By most accounts, she was an ordinary woman….

via A WOMAN NAMED ELLEN — The Neighborhood

Why I Am Suing the Government

Social Media Collective

(or: I write scripts, bots, and scrapers that collect online data)

I never thought that I would sue the government. The papers went in on Wednesday, but the whole situation still seems unreal. I’m a professor at the University of Michigan and a social scientist who studies the Internet, and I ran afoul of what some have called the most hated law on the Internet.

Others call it the law that killed Aaron Swartz. It’s more formally known as the Computer Fraud and Abuse Act (CFAA), the dangerously vague federal anti-hacking law. The CFAA is so broad, you might have broken it. The CFAA has been used to indict a MySpace user for adding false information to her profile, to convict a non-programmer of “hacking,” to convict an IT administrator of deleting files he was authorized to access, and to send a dozen FBI agents to…

View original post 835 more words