Category: show on landing page

[ALEX] op-ed: Personalization algorithms and elections

fbtrexOn February 7, 2019 the Internet Policy Review published an op-ed by Stefania Milan and Claudio Agosti. We reflect on personalization algorithms and elections, and share some ideas about algorithmic sovereignty and literacy. Thanks to Frédéric Dubois for the invitation.

“Personalisation algorithms allow platforms to carefully target web content to the tastes and interests of their users. They are at the core of social media platforms, dating apps, shopping and news sites. They make us see the world as we want to see it. By forging a specific reality for each user, they silently and subtly shape customised “information diets”, including around our voting preferences. We still remember Facebook’s CEO Mark Zuckerberg testifying before the US Congress (in April 2018) about the many vulnerabilities of his platform during election campaigns. With the elections for the European Parliament scheduled for May 2019, it is about time to look at our information diets and take seriously the role of platforms in shaping our worldviews. But how? Personalisation algorithms are kept a closely guarded secret by social media platform companies. The few experiments auditing these algorithms rely on data provided by platform companies themselves. Researchers are sometimes subject to legal challenges by social media companies who accuse them of violating the Terms of Services of their utility. As we speak, technological fencing-offs are emerging as the newest challenge to third-party accountability. Generally, auditing algorithms fail to involve ordinary users, missing out on a crucial opportunity for awareness raising and behavioural change.

The Algorithms Exposed (ALEX) project1, funded by a Proof of Concept grant of the European Research Council, intervenes in this space by promoting an approach to algorithms auditing that empowers and educates users. ALEX stabilises and expands the functionalities of a browser extension – fbtrex – an original idea of lead developer Claudio Agosti. Analysing the outcomes of Facebook’s news feed algorithm, our software enables users to monitor their own social media consumption, and to volunteer their data for scientific or advocacy projects of their choosing. It also empowers advanced users, including researchers and journalists, to produce sophisticated investigations of algorithmic biases. Taking Facebook and the forthcoming EU elections as a test case, ALEX unmasks the functioning of personalisation algorithms on social media platforms.”

Continue reading in the website of the Internet Policy Review.

[bigdatasur] Jitihada La Ughaibu Afrika

By Duncan KinuthiaFord/Media Democracy Fund Tech Exchange Fellow at Research ICT Africa

>>> if your Swahili is not good enough, check out the English translation! <<<

Watumizi wa mtandao Afrika wanazidi kuongezeka katika kusaka ughaibu kwenye mtandao. Hii imehamasishwa na ripoti za hivi majuzi za ukiukaji wa uaminifu na faragha wa data katika mitandao ya kijamii. Hii pamoja na kuvuka mpaka katika usalama na udhibiti wa serikali imesababisha kuongezeka kwa vifaa vya ughaibu wa data. Ama kweli, watumizi wanafanya hima ili kuweza kupata habari safi, usalama na faragha mtandaoni. Kwa vile muundo wa mtandao hauruhusu ughaibu kamili, waeza kupata ughaibu huu kwa kutoa majina ama habari itakayokubainisha na kwa kutumia teknolojia fiche katika habari za watumizi wa wavuti. Utumizi wa VPN (Virtual Private Network) imeongezeka kwa watumizi wa wavuti duniani pamoja na makampuni kwa ajili ya ongezeko la teknolojia fiche zipatikanazo kwenye VPN katika usambazaji wa data kwenye mtandao ulio na upungufu wa usalama. Hii imesababisha ukuaji wa maonyesho wa soko la VPN muongo uliopita na pia MarketWatch kuripoti asilimia kumi na nane katika kiwango cha ukuaji kila mwaka. Miunganisho ya VPN imeimarishwa kwa kutumia njia ya teknolojia fiche ikifuatiliwa na uthibitisho ya lazima kwa mtumiaji ili aweze kupata kuunganishwa kwa hiyo VPN.

Afrika imekabiliwa na mlipuko wa matumizi wa simu zilizo na upatikanaji wa mkondoni muongo uliopita ambao umesababisha kuongezeka kwa watumiaji wa mtandao barani Afrika. Mlipuko huu umeleta mageuzi bora ya kiuchumi, kisiasa na kijamii. Mojawapo ikiwa watumiaji zaidi wakiingia katika mitandao ya kijamii ili kuweza kuendelea kuwasiliana na familia na marafiki na hata kueneza shughuli za kiuchumi. Biashara ya wavuti pia imeenea Afrika ikichochewa na ukuaji wa matumizi ya huduma za fedha za simu kote barani. Ilhali utumizi wa mtandao umeleta faida kikandani kuna udhibiti, kuzimisha na kodi ya mitandao ya kijamii pia imezidi. Matendo haya ya kiukaji yaweza kupunguza faida za digitization.

Kwa watumizi wa mtandao wengi Afrika mitandao ya kijamii ndiyo mtandao halisi kwao na gharama kubwa ya bandwidth ndiyo kikwazo kikuu cha utumizi wake. Mfano ni watumizi wa Uganda ambao walihangaishwa na ushuru liyofanywa rasmi kuanzia tarehe moja Julai mwaka wa 2018. Sheria, iliyopitishwa na bunge la Uganda inatia kodi ya shilingi mia mbili za Uganda ($ 0.05) kwa matumizi ya mitandao ya kijamii kila siku. Hii ni sawa na dola 19 ($19) kwa kila mwaka na pamoja na gharama kubwa za bandwidth inazuia sana matumuzi ya mitandao ya kijamii, kutokana na kwamba jumla ya bidhaa za nyumbani kwa kila mtu ilikuwa dola mia sita na nne ($ 604) tu mwaka 2017. Ili kuepuka ushuru wa matumizi ya mitandao ya kijamii, watu wanatafuta njia za kuihepa. Utafiti uliofanywa juu ya kodi ya mitandao ya kijamii nchini Uganda umebaini kwamba asilimia hamsini na saba ya watumiaji waligeuka kwa huduma za VPN ili kuepuka kodi iliyolazimishiwa kwao.

Hata hivyo, hizi sio sababu za pekee ambazo watumiaji wa Afrika wanatumia VPN. Kwa mfano, Kenya, watumiaji wengi wa mtandao hutumia VPN na kuhadaa DNS kuepukana na vitengo vya Geo, ambapo maudhui mengi ya elimu na burudani haipatikani nchini kwa sababu ya leseni, hati miliki na ukosefu wa soko kubwa ili kuhakikisha kurudi kwa uwekezaji, kwa kuwa mtandao bado haujapenya vizuri barani. Huduma hizo zinajumuisha Spotify, vipindi na sinema nyingi kwenye Netflix, muziki wa YouTube, Google Music, Google Play Books, Pandora na huduma zinginezo. Watumiaji wa mtandao wameamua kutumia VPN na kuhadaa DNS ili waweze kupata huduma hizi.

Katika nchi nyingine, matumizi ya mtandao ya kijamii yanaonekana kuwa tishio kwa uanzishwaji, na serikali zimeweka masharti ya kisiasa ya mtandao. Vikwazo vya mtandao vinatumiwa kukabiliana na matumizi ya huduma za VPN ili kupata maudhui yaliyolengwa, kama vile matumizi ya uhusiano wa VPN juu ya mtandao inategemea uhusiano wa mtandao. Idadi kubwa ya vikwazo vya mtandao vilivyoripotiwa katika nchi za Afrika hufanyika wakati wa uchaguzi, juu ya madai ya kudhibiti uenezi wa habari bandia. Wananchi wa Jamhuri ya Kidemokrasia ya Kongo ni waathirika wa hivi karibuni kwa hili, baada ya kuzimwa kamili kwa mtandao wakati wa uchaguzi wa Desemba 30. Nchi nyingine za Kiafrika ambazo zimekuwa na vikwazo vya mtandao ni kama Ethiopia, Cameroon, Gambia na Gabon.

Ufahamu wa kikanda juu ya hatari za ufunuzi wa habari na ukiukaji kwenye wavuti ni duni, kwani wafrika wengi hawajajaliwa kutumia mtandao. Kwa kuongeza, wale wanaotumia mtandao hawajui vitisho vya wavuti. Ripoti ya Utoaji wa Usalama wa Afrika ya 2016 ilibainisha kuwa asilimia hamsini ya waliohojiwa hawakupewa mafunzo ya usalama wa cyber. Hii imechangia kuongezeka kwa gharama za makadirio ya uhalifu wa wavuti, Nigeria ikiwa na gharama kubwa zaidi ya $550 milioni. Hata hivyo, kama matumizi ya mtandao yanakua Afrika, haja ya kuhakikisha usalama wa habari kulinda utambulisho wa watu na matumizi ya bure ya mtandao, huja kama jambo muhimu. Kuelewa mbinu za sasa za ughaibu wa data, zana na mazoea kote kandani na jinsi hatua za usalama za habari zinazotumiwa na watumiaji wa Afrika ni muhimu kwa kuinua ufahamu, na hivyo kuleta ushahidi kwenye mjadala wa sasa wa sera juu ya faragha na usalama mtandaoni kutoka kwa mtazamo wa Afrika.

Mwaka mmoja ujao, nitafanya kazi chini ya Media Democracy Fund Tech Exchange Program, na Research ICT Africa, katika utafiti juu ya udhibiti wa habari katika Afrika Mashariki, ikiwa ni pamoja na matumizi ya mbinu za ughaibu wa data na ufanisi wa DNS, lengo muhimu likiwa kuhusu matumizi ya VPN. Research ICT Afrika ni kundi la wataalam wanaojadili sera ya kikanda ya ICT na wanafanya utafiti mbalimbali juu ya utawala digital, sera na kanuni zinazowezesha sera zilizoarifiwa kwa ajili ya upatikanaji bora, matumizi ya teknolojia ya digital kwa maendeleo ya kiuchumi na kijamii Afrika kwa kutumia ushahidi hakika.

Lengo kuu la mradi wangu wa utafiti ni kuangazia mazoezi ya kutumia VPN kama chombo cha ughaibu wa data katika Afrika Mashariki, kwa watumiaji binafsi na mashirika yasiyo ya faida (NPO).

Miongozo itakayoongoza uchunguzi na utafiti wangu ni haya:

  1. Ni sababu gani kuu za kutumia VPN kama mbinu ya ughaibu wa data katika Afrika Mashariki?
  2. Ni nani watumiaji wakuu wa VPN katika Afrika Mashariki na ni nini mwenendo katika   vikundi tofauti vya watumiaji kwa umri na jinsia?
  3. Kwa nini watumiaji wa mtandao na NPO hutumia VPN katika Afrika Mashariki?
  4. VPN imetumiwa wapi zaidi?
  5. Na VPN ilitumiwa lini katika Afrika Mashariki?
  6. VPN na zana zingine za ughaibu wa data zimetumiwaje ili kuhakikisha usalama na ufaragha wa habari Afrika Mashariki?

Utafiti juu ya matumizi ya zana za ughaibu wa data katika Afrika ya Mashariki utakuwa muhimu kwa kueneza ufahamu kwa wanaharakati wa data. Mkusanyiko wa data juu ya mbinu mbali mbali zinazotumiwa na watumiaji wa mtandao katika Afrika ya Mashariki kufikia kutokujulikana kwa mtandao zitatiweka dhahiri mbinu za ubunifu ambazo watu hutumia kuhepa vikwazo vilivyowekwa na mashirika makubwa na serikali.

Baadhi ya matokeo ya uchunguzi yataonyesha mienendo ya matumizi ya zana za uonyesho wa data wakati wa vipindi vya uchaguzi wa nchi zilizoathiriwa na kuzimwa kwa mitandao ya kijamii. Utafiti huo utafafanua pia jinsi makundi mbalimbali ya watumiaji hutumia VPN ili kuwezesha upatikanaji wa burudani ya jiji iliyozuiwa pamoja na maudhui ya elimu au usalama wa data kutokana na vipengele vya teknolojia fiche.

Kwa suala la mapendekezo ya sera, utafiti huu utasaidia kuelewa wa dhana ya data kutoka kwa mtazamo wa Kiafrika na utajulisha mjadala wa sera za kikanda na kimataifa juu ya faragha, usalama na usalama mtandaoni katika Afrika. Utafiti huo pia utatoa mapendekezo kwa watumiaji wa internet wa Afrika juu ya jinsi ya kuwa salama mtandaoni kwa njia ya uhamisho wa data kote kanda.

 

 

Student/volunteer needed for development of the plugin Facebook.tracking.exposed

As part of the spin-off project on algorithmic personalization, the DATACTIVE team is looking for an enthusiastic volunteer who would like to engage in qualitative market research for the development of the facebook.tracking.exposed tool, a browser plugin that allows users to “re-appropriate” Facebook timeline data for research purposes. The position starts halfway February for two days a week (flexible) for a period of one month with possibility of extension should both parties be interested. If interested, please contact project manager Jeroen de Vos before Tuesday February 5th, he can be reached at jeroen@data-activism.net

More information: DATACTIVE, ALEX (coming soon: algorithms.exposed)

Stefania at CPDP & PrivacyCamp in Brussels

Stefania will be in Brussels to take part in the Privacy Camp and in the Computer, Privacy and Data Protection (CPDP) conference, in Brussels at the end of January.

At Privacy Camp, on January 29, Stefania will join Mirko Tobias Schäfer (Utrecht Data School), Minna Ruckenstein (University of Helsinki) and DATACTIVE former postdoc Jonathan Gray (now at Kings’ College, London) for a discussion on Reimagining Data Futures: Data and Agency.

At CPDP, Stefania will join a panel on “Data is (Political) Power!” (Wednesday 4pm) organised by IViR & ASCoR (UvA), with Daniel Kreiss (University of North Carolina at Chapel Hill), Julia Reda, EU MEP, and Jennifer Childs (ICO). On Friday, she will also contribute to an amazing session of the artistic program, entitled Micro-Targeting and Tactical Fiction, featuring Manu Luksch (!).

Check out the CPDP program!

 

Good Data Book launch on January 24

For the past few months we worked in a new collective DATACTIVE publication: Data for the Social Good: Toward a data-activist research agenda. It will be one of the chapters in the forthcoming Good Data book, edited by Angela Daly, Kate Devitt and Monique Mann, and published by the Institute of Network Cultures in Amsterdam within their “Theory on Demand” series.

Our text builds upon the Data for the Social Good two-day focused encounter we organised in November 2017 (read the report here). During the first day of that event last year we discussed with Charlotte Ryan, Lorenzo Pezanni, Jeff Deutsch and Nico Para about the ways in which research and activism intersect in projects which rely on data. On the second day, we were joined by a diverse group of researchers and activists for a workshop exploring what a data-activist agenda would look like (seriously, read the report here!).

In Data for the Social Good: Toward a data-activist research agenda, we took that conversation and expanded it two fronts. First, we grounded theoretically our take on what it means to be a (data) activist, which implied clarifying what data activism means in the first place, as well as briefly revising the origin and evolution of engaged research. Secondly,  we reflected more deeply upon the ethics of collaborative investigations, paying particular attention to the power relations between the actors involved throughout the process.

The Good Data book will be launched in Amsterdam at 5pm of the 24th of January, at the cultural centre Spui25. It includes 20 different chapters exploring what good data practices are, from manifestos to smart city reflections (and nope, not a single reference to blockchain).  Both the editors and DATACTIVE will be around to discuss more about how to use data for good rather than evil, so join us if you’re interested in knowing more (or getting the book)!

Cheers!

DATACTIVE

 

 

ALEX @DMI winterschool

Between the 7th and 11th of January, Claudio Agosti, Davide Beraldo and Jeroen de Vos from the ALEX / DATACTIVE team took part in the annual Digital Methods Initiative’s Winter School. This was the perfect occasion to kick-off a project related to the development and application of facebook.tracking.exposed, a browser extension developed in order to expose the functioning of the secretive Facebook’s News Feed algorithm and adopted by the Algorithms Exposed (ALEX) project.

The ALEX project pitch collected quite some interest, with about 15 people coming together for a hectic as much as a fun week of collaborative thinking, experimenting and analyzing. Given the many possibilities of exploring and the variety of available skills, the group split into 2 subgroups: one has been busy with the creation of brand new ‘bots’, with the experimental interest in assessing the role of emotional engagement and friendship making in the Facebook timeline algorithm; the second group has been working on an existing dataset related to last year’s Italian national elections.

You can take a deeper look at what has been done and the insights that have been collected in the Winter School’s project wiki (TBA) and in the final presentation. Here is a bullet-point summary of the main findings:
• a bot’s life is a dangerous life… you gotta be smart not to be Facebook-killed
• being a bot is not that boring though… many new bot friends are ready to connect to you
• love wins over hate… consistently love-reacting posts seems to trigger more content on the timeline than consistently “angry”/negative reactions
• tell me what you liked, I tell you what you’ll see… selectively liking posts from different political orientations will affect the political issues you will see on your future timeline
• and lastly, and this is controversial… centre-left wing bots are more prone to be exposed to controversial content than far-right wing bots.

Many thanks to the organizers of the Winter School and, of course, to all those who contributed to the Algorithms Exposed group: Iain Emsley, Fatma Yalgin, Hannah Vischer, Victor Pak, Claudio Agosti, Mathilde Simon, Victor Bouwmeester, Yao Chen, Sophia Melanson, Hanna Jemmer, Patrick Kapsch, Giovanni Rossetti, Davide Beraldo, Giulia Corona, Leonardo Sanna, Jeroen de Vos.

Announcing the Magma project

By Vasilis Ververis, DATACTIVE

Magma aims to build a scalable, reproducible, standard methodology on measuring, documenting and circumventing internet censorship, information controls, internet blackouts and surveillance in a way that will be streamlined and used in practice by researchers, front-line activists, field-workers, human rights defenders, organizations and journalists.

In recent years, a number of research fellows, journalists, human rights activists, lawyers as well as a larger research community, have been working in high-risk contexts, which creates the need to consider their qualitative and quantitative research data as highly sensitive. Albeit their competitiveness and high qualification in their respective areas (social and political science, usability, law, political economy analysis), they can rarely claim to have a specific expertise or extensive experience when it comes to networks services and systems, telecommunication infrastructure, applied data analysis of network measurements, internet censorship, surveillance and information controls.

Ideally, researchers working with various network measurement tools and frameworks such as the Open Observatory of Network Interference (OONI), should have qualified technical help and assistance, thus enabling them to develop appropriate testing methodologies, suiting exactly their research environment and needs.

Magma aims to build a research framework for people working on information controls and network measurements, facilitating their working process in numerous ways. As such, this framework will enable them to properly structure an activity plan, make informed choices regarding the required tools (including ethical and security aspects) and analyze the data produced by such tools.

Through Magma, we wish to provide our expertise and experience in network measurements, internet censorship research, assessment of ISP network, surveillance probing and data analysis in order to:

  • Asses the risks by providing, implementing and maintaining technologies demanded by researchers on front-lines and areas where the need of operational security, anti-surveillance and censorship circumvention is of paramount importance.
  • Provide tailored technical assistance, developing at the same time appropriate testing methodology for network measurements, evaluation and analysis of data and reports that correspond to the respective research questions.
  • On a long-term basis, build a scalable and reproducible methodology for collecting, evaluating and analyzing data and reports’ self-defense for front-line researchers, front-line activists, field-workers, human rights defenders, organizations and journalists, by keeping exact documentation.

Below, we list some examples of potential future research around internet censorship, information controls and surveillance, mainly based on conducting networks measurements and analyzing its results:

Egypt: Media censorship, Tor interference, HTTPS throttling and ads injections?

A study on Tor network and media websites blockages, network bandwidth throttling and malicious network packet injections that contained malware and advertising content.

OONI Data Reveals How WhatsApp Was Blocked (Again) in Brazil

A study to determine how WhatsApp has been blocked after a judge’s court order all over the country of Brazil.

Understanding Internet Censorship Policy: The Case of Greece

An extensive large scale research analyzing the policies and techniques used to block content deemed illegal by a state identifying transparency problems, collateral damage and the implications of over or under blocking.

Identifying cases of DNS misconfiguration: Not quite censorship

A study on a non-malicious technical issue that leads to the interference and non-accessibility of a regional news media outlet throughout several different networks and countries.

To this respect, we would like to hear from all of you who are interested in researching information controls and internet censorship, and are intrigued to better understand how to work with network measurements and analyze data from various data sources and OONI reports.

We wanted to keep this post as concrete and terse as possible to encourage both technical and non-technical entities and individuals to get in touch with us, even if they are currently engaged in an undergoing project. The results of this collaboration will help form a complete guideline handbook expressed by the needs of the communities that work, or conduct research, in this field.

Please use any of these communications channels to get in touch with us.

 

Vasilis Ververis is a research associate with DATACTIVE and a practitioner of the principles ~ undo / rebuild ~ the current centralization model of the internet. Their research deals with internet censorship and investigation of collateral damage via information controls and surveillance. Some recent affiliations: Humboldt-Universität zu Berlin, Germany; Universidade Estadual do Piaui, Brazil; University Institute of Lisbon, Portugal.

 

This post is co-published with the Magma Project

new article: Playing with data and its consequences out in First Monday

The January 2019 issue of First Monday includes an article by research associate Miren Gutierrez and PI Stefania Milan on the consequences of engagement with data. The article is entitled “Playing with data and its consequences” and is the lead article of the issue. Check it out!

Abstract. The fundamental paradigm shift brought about by datafication alters how people participate as citizens on a daily basis. “Big data” has come to constitute a new terrain of engagement, which brings organized collective action, communicative practices and data infrastructure into a fruitful dialogue. While scholarship is progressively acknowledging the emergence of bottom-up data practices, to date no research has explored the influence of these practices on the activists themselves. Leveraging the disciplines of critical data and social movement studies, this paper explores “proactive data activism”, using, producing and/or appropriating data for social change, and examines its biographical, political, tactical and epistemological consequences. Approaching engagement with data as practice, this study focuses on the social contexts in which data are produced, consumed and circulated, and analyzes how tactics, skills and emotions of individuals evolve in interplay with data. Through content and co-occurrence analysis of semi-structured practitioner interviews (N=20), the article shows how the employment of data and data infrastructure in activism fundamentally transforms the way activists go about changing the world.

Citation: Gutierrez, Milan and Stefania Milan (2019). Playing with data and its consequences, First Monday, Volume 24, Number 1 – 7 January 2019, https://firstmonday.org/ojs/index.php/fm/article/view/9554/7716
doi: http://dx.doi.org/10.5210/fm.v24i1.9554

 

A year in review

2018 has been a good year for DATACTIVE. We take the opportunity of the turn of the year to review what we accomplished and what remains to do.

We advanced with data collection, and are almost done. Just to mention one, we are close to 200 interviews, and the material is extremely rich.

We organized two exciting events, the July workshop and the Big Data from the South, and a hackaton.

Collectively we delivered over 40 talks.
We published a dozen between papers and book chapters, including an article for Policy & Internet, two for the International Journal of Communication (here and here), three contributions in a special issue of XRDS on anonymity. We released a special issue on data activism of the peer-reviewed journal Krisis: Journal for Contemporary Philosophy. Many more contributions are in print. In January alone, an article on the consequences of engaging with data will appear on First Monday; a collective book chapter will be released in the context of the collection on Good Data; the special issue of the journal Policy & Internet on internet architecture & human rights will see the light of day.
The DATACTIVE blog, the critical communities debate and the Big data from the South blog are thriving; our work was mentioned in several media outlets in a variety of idioms.
In July we were awarded a Proof of Concept grant of the ERC to work on the Algorithms Exposed project (ALEX). Stay tuned for further developments, including a brand-new website which will soon become available at the URL algorithms.exposed.
And most importantly, we continued learning from each other and from the many activists we encountered during fieldwork, and we continued experimenting with a different way of doing and being academia (among others, see  here and here).

With less than two years to the end of the grant, we will now dedicate ourselves primarily to data analysis and writing. ALEX will keep some of us busy, and will allow us to expand our team hiring a couple of developers and collaborating with NGOs. To start with, next week we will seize the opportunity of the forthcoming of the Digital Methods Winter School to advance with software development. We can anticipate we will use the forthcoming EU Parliament election as one of our test cases, so if you are interested in collaborating to a research on the effects of algorithmic personalisation please get in touch.

Yours truly, the DATACTIVE team