ICT4D lecture: HCI4D Human-Computer Interaction for Development
Download as PPTX, PDF2 likes2,680 views
The document discusses the intersection of Human-Computer Interaction for Development (HCI4D) with cross-cultural contexts, highlighting the importance of designing technology for low-literate users in developing regions. It presents various methodologies and case studies, including the use of voice-based user interfaces and localized designs to enhance accessibility and usability in diverse environments. The overall aim is to leverage information and communication technologies (ICT) to improve the lives and livelihoods of people in these challenging contexts.
![“Human-Computer Interaction for
Development” (HCI4D).
To study
“[…] how interactive products, applications, and
systems can be appropriately designed to
both address the distinctive needs of users in
developing regions, and to cope with the
difficult infrastructural contexts where these
technologies must work.”
-Ho et al.](https://image.slidesharecdn.com/hci4d-140414090837-phpapp01/85/ICT4D-lecture-HCI4D-Human-Computer-Interaction-for-Development-5-320.jpg)
![Also..
“[…] we do not seek merely to understand how
humans and ICTs interact in developing
regions, but to apply this understanding to
adapt the interactive behavior of ICTs in these
contexts, to shape new and more appropriate
forms of ICTs, and to devise human centered
approaches to designing ICTs that can be used
by people to improve lives, livelihoods, and
freedoms.”](https://image.slidesharecdn.com/hci4d-140414090837-phpapp01/85/ICT4D-lecture-HCI4D-Human-Computer-Interaction-for-Development-6-320.jpg)
![In multicultural environments it is even more
important [to] consider how our
understanding of the complex dialectic
between culture, economy and technological
innovation influences our ability to empower
our people. (p. 4)
Hugo, J. (2002). HCI and multiculturalism in southern Africa.
Proceedings of the CHI 2002 Development Consortium.](https://image.slidesharecdn.com/hci4d-140414090837-phpapp01/85/ICT4D-lecture-HCI4D-Human-Computer-Interaction-for-Development-7-320.jpg)
ICT4D lecture: HCI4D Human-Computer Interaction for Development
- 1. HCI4D ICT4D lecture 14/4/2014 - Victor de Boer with slides from Stephane Boyera
- 2. Outline • General introduction to HCI4D – Cross-cultural HCI – Low literate-interfaces • Case: Voice access to Web content • Project updates • Thursday: VXML tutorial and assignment setup
- 3. One minute paper • What is the most important idea I got from Ho et al. (or two or three)
- 5. “Human-Computer Interaction for Development” (HCI4D). To study “[…] how interactive products, applications, and systems can be appropriately designed to both address the distinctive needs of users in developing regions, and to cope with the difficult infrastructural contexts where these technologies must work.” -Ho et al.
- 6. Also.. “[…] we do not seek merely to understand how humans and ICTs interact in developing regions, but to apply this understanding to adapt the interactive behavior of ICTs in these contexts, to shape new and more appropriate forms of ICTs, and to devise human centered approaches to designing ICTs that can be used by people to improve lives, livelihoods, and freedoms.”
- 7. In multicultural environments it is even more important [to] consider how our understanding of the complex dialectic between culture, economy and technological innovation influences our ability to empower our people. (p. 4) Hugo, J. (2002). HCI and multiculturalism in southern Africa. Proceedings of the CHI 2002 Development Consortium.
- 8. Cross-cultural HCI • Study of Human-Computer Interaction across cultural differences • Cf. Cross-cultural Interface design – Internationalization (I18n) – Localization (I10n) – “Communitization” – Hugo
- 9. Localization • Translation • Adapt and Adjust • Details – Dates: Month, day, year vs. day, month, year. – Time: 12-hour vs. 24-hour time. – Color: Avoid local color sensitivities. – Currency: Pay attention to conversions and formats. – Phone Numbers: Formats are different around the world. – National Holidays: Holidays are country and region specific. – Geographic Examples: Keep it relevant for your audience. – Website Language Codes: ISO codes are important to know. http://mashable.com/2011/12/22/website-localization/
- 11. Example: individualism Italy:76, Venezuela: 12 https://www.humanfactors.com/downloads/apr13.asp
- 12. A lot of work in • Cross cultural HCI • Localization • Internationalization International development is almost never a goal in cross-cultural HCI.
- 13. User-centered design methodology • Participatory approach (cf. International Development) – Political environment, ethics, cultural differences • Living labs-like – Early prototyping (use comics with children) • Action research and the many responsabilities of the HCI4D researcher Serrewet.com
- 14. Information access for the rural poor
- 15. Copyright © 2013. SBC4D – http://wwww.sbc4d.com. All rights reserved 15 The promises of ICTD o WSIS 2003/2005: Technology is a development tool: o Overcoming infrastructure challenges o Education o Healthcare o Livelihood o etc. o Leveraging communication independently of physical/geographical barriers o Improving transparency, accountability, efficiency of governments
- 16. Copyright © 2013. SBC4D – http://wwww.sbc4d.com. All rights reserved 16 Web Access: Current State
- 17. Copyright © 2013. SBC4D – http://wwww.sbc4d.com. All rights reserved 17 Information Appliances: Mobile phones
- 18. Copyright © 2013. SBC4D – http://wwww.sbc4d.com. All rights reserved 18 Information Appliances: Radio
- 19. Many examples with SMS Medic.frontlinesms.com
- 21. Low-literate users • One of the grand challenges (Ho et al.) • Especially prevalent among the rural poor • Cf. Accessibility here – Web design http://www.w3.org/TR/WCAG20/
- 22. Varying levels • Also: numeracy • Always base yourself on information as specific as possible
- 23. Quiz • Literacy in Zimbabwe A. 0-30% B. 30-50% C. 50-70% D. 70-100%
- 24. Quiz • Literacy in Zimbabwe A. 0-30% B. 30-50% C. 50-70% D. 70-100% <- 91%
- 25. Quiz • Literacy in Mali A. 0-30% B. 30-50% C. 50-70% D. 70-100%
- 26. Quiz • Literacy in Mali A. 0-30% <- 27.7% (20% among women) B. 30-50% C. 50-70% D. 70-100%
- 27. Varies greatly in continent Data.worldbank.com
- 28. Icon-based interfaces • Example: Indrani Medhi – Correlation between ease of using hierarchical menu’s (even without text) and literacy
- 29. Interaction design for low- literacy communities
- 30. Feature phones
- 32. Icon-based interaction NCR ATM interface for illiterate 'grammar' - ISOTYPE by Otto Neurath available at http://imaginarymuseum.org/MHV/PZImhv/NeurathPictureLanguage.html
- 33. Voice-based user interfaces • Not exclusive to ICT4D (cf Siri, In-car systems) • Very suitable for oral cultures
- 35. Voice menus for interactive information access
- 36. Voice menu for Banana service: P. Nasfors. Efficient Voice Information Services for Developing Countries, Master Thesis, Department of Information technology, Uppsala University, Sweden, 2007.
- 37. VUI design guidelines D. Chhetri MsC Thesis “Voice User Interface Design for the m-Event Organizer”
- 38. VUI design guidelines • Input requirements – DTMF* rather than voice input – Local phone nrs • Structure requirements – Providing an overview of the page – Eliminate repetitions – Immediate Feedback – Error recovery • Output requirements – System voice: female (75%) and human (67%) * probably case and culture-specific – Nonverbal sounds R Nieland Msc Thesis (tentative): Talking to Linked Data (*)Dual-tone multi-frequency signaling
- 39. ‘Small Languages’ • Large language populations: English, German, Chinese, Portuguese, Dutch… – Large number of textual corpora available, great economic value to localization • Small languages: Frisian, Bambara (3M speakers), – Small number of textual corpora available, some economic value to localization • Very small languages: Bomu (30K speakers) – No textual corpora available, very low economic value to localization
- 40. Computer-literacy • Digital divide – Menu? – Home? – Save? • Cf. usability difficulties that South African subjects encounter with hierarchical information structures to the non-tree-like schema that the subjects use to conceptualize their “family trees.” Walton, M., & Vukovic, V. (2003). Cultures, literacy, and the Web: Dimensions of information “scent.” interactions, 10(2), 64–71.
- 41. Voice-based Web access in Africa
- 43. Can the Web (be made to) mean something for knowledge sharing even under very constraining conditions? No internet, no computer, no electricity Multitude of languages, levels of literacy
- 44. A window of opportunity…
- 45. •No. 1 source of information •Interactive radio programs •Huge listening base Another window: rural community radios
- 46. • Integrate local community radios and mobile ICT for knowledge sharing • Better support and integrate local languages in voice-based services – Development of appropriate speech elements (text-to-speech and Speech recognition) • Develop a free and open source toolbox for local developers. • Investigate self-sustainability – Develop appropriate business models – In collaboration with local communities.
- 47. June 2013Jan 2010 First cycle: Knowledge sharing for regreening Second cycle: Use Case elicitation and selection RadioMarché: Pilot System development (Malian French) RM: 3rd version (Bambara,Bomu ) Jan 2011 Jan 2012 Jan 2013 First field trip: Burkina Faso Second field trip Mali, Burkina Faso, Ghana RM: 4th version Third field trip Mali Fourth field trip Mali Fifth field trip Mali Foroba Blon: pilot version FB: 2nd version (Bambara,Bomu) Third cycle: Requirements elicitation and validation Business Model integration Sixth planned field trip Mali Fourth cycle Use case and Requirements validation Fifth cycle: Use case and Requirements validation Sept 2009 Cycle 0 Preliminary visit BF sept 2009 RM: 2nd version (production version) FB: 3rd version (Bambara,Bomu)
- 48. Three voice systems … Market Information Citizen Journalism Event Organiser
- 50. Foroba Blon use case • Citizen journalism platform – using the same technology as RadioMarché • Multiple use cases – Letters to editor – Trusted users ‘journalists’ • Funded by IPI news innovation contest
- 51. Foroba Blon
- 54. Adopted by local radio stations Positive feedback so far. • 250 messages left Messages about deaths, missing animals Problems with registration Voice Interface difficulties Costs New business models: ‘advertiser’ pays through credit Voice Twitter? Al Jazeera used Foroba Blon for Ghanaian (2012) and Kenyan (2013) elections coverage. Foroba Blon Results
- 57. Event Organising
- 58. Event Organizer • Event organizer for Sahel Eco • User profile – Name, phone number, language(s) • Call everyone and note responses • “Tabale”: the sacred drum
- 59. The Tabale Platform NGO staff Malian French Bomu Bambara Record multiple messages in different languages yes no don’t know
- 60. VUI design (three languages)
- 63. M-agro Use Case Context: Regreening in Africa
- 65. Communiqué
- 66. Local market data Communiqué GSM/Voice interface Web Interface Text-To-Speech Community radioSahel Eco operative Buyers
- 67. SMS Radio 1: French and Bambara Radio 2: French and Bomu
- 68. SMS Radio 1: French and Bambara Radio 2: French and Bomu
- 69. Web frontend for Market data management and communiqué generation RadioMarche.com Iterative, user-centered design. Uses same information architecture as legacy system (cf. Communique sheets)
- 70. Voice channel: Multiple solutions Netbook running Prophecy SIP over Ethernet Netbook HTTP Officeroute running Asterix Orange Emerginov Platform Netbook running Prophecy + Asterix Bluetooth Mobile phone
- 71. Audio communiqué • Audio communiqué generated by slot-and- filler Text-to-speech system – concatenating prerecorded phrases – Each radio host uses own voice (!) • Malian French • Bambara • Bomu Radio Segou Radio Mountian
- 72. “Slot and Filler” Text-to-Speech English: Bambara: 15 liters of offered by Zakari Diarra 15_ba.wav L_ba.wav Of_ba.wav Spoken Language Elements Repository honey
- 73. VoiceXML <?xml version="1.0" encoding="ISO-8859-1"?> <vxml version="2.0" lang="en"> <form> <prompt bargein="false"> Welcome to RadioMarche! <audio src=“audio/communique_1_bambara.wav"/> </prompt> <option dtmf="1" value=“1">Press one for X</option> <option dtmf=“2" value=“2">Press two for Y</option> ... </vxml> DTMF = Dual-tone multi-frequency signaling
- 75. Deployment
- 76. Evaluation • Limited usage due to: – Political situation – Ecological situation – Social barriers – Technical issues • More languages: Bambara and Bomu – Bomu way of counting • More usage • In a way, we are “too successful”: Too much demand – not serving the value chain – Should we include this in the application? – Cf. Marktplaats?
- 77. HCI4LD4D • Linked Data for sharing information across contexts, locations – Light-weight integration of data • Language-agnostic! • Interface-agnostic!
- 80. Linked Market Data • 90 offerings • 19 contacts • Links to – Data • DBPedia • GeoNames • Agrovoc – Vocabularies • Foaf • GoodRelations
- 81. Sharing across regions Local market data RadioMarché market information system Farmers (producers) Buyers (consumers) Email GSM/VoiceWeb SMS Data / communique platform Local radio RadioMarché in second region Local market data Data / communique layer Farmers (producers) Buyers (consumers) Email GSM/VoiceWeb SMS Interface handler layer Local radio
- 82. DBpedia GeoNames Agrovoc Inst. of Development Studies LOD Citizen Journalism data Sahel Pluvial data SemanticXO RadioMarché Linked market data Linked Data for Development (LD4D) IATI data
- 85. rm:offering0001 rm:shea_butter rm:1000 rdfs:label rdfs:label “Amande de Karité”@fr “Shea Nuts”@en speakle:voicelabel_ba rm:audio_shea_nl.wav rm:audio_shea_ba.wav speakle:voicelabel_nl rdfs:label “1000” speakle:voicelabel_ba rm:audio_1000_nl.wav rm:audio_1000_ba.wav speakle:voicelabel_nl rm:Mazankuy_Diarra rm:kilo rdfs:label “kilo”@en speakle:voicelabel_ba rm:audio_kilo_nl.wav rm:audio_kilo_ba.wav speakle:voicelabel_nl rm:has_contact Speakle voice labels
- 90. <vxml version = "2.1" > <property name="inputmodes" value="dtmf" /> <form id="result"> <block> <prompt> These are the top three current offerings for Tamarin <break time="0.5s"/> <audio src="http://.../AndiaKalakodio_en.wav"/> offers <audio src="http://.../450_00_en.wav"/> <audio src="http://.../Kg_en.wav"/> for <audio src="http://.../250_00_en.wav"/> <audio src="http://.../CFA_en.wav"/> <break time="0.5s"/> ... <break time="0.5s"/> </prompt> <goto next="mytest.xml"/> </block> </form> </vxml> Voice browser Tel: +31208080855 Skype: +990009369996162208
- 91. To conclude… • Linked Market Data – Locally created – Linked Data make re-use possible (NGO, others) – LD voice labels • Can be (re)used to develop voice applications with this data • To go beyond proof-of-concept – More localization needed – Local hardware/services (Emerginov / OfficeRoute) – User testing – More sophisticated translations (VoiceSPARQL)
- 93. Web for ALL. Using voice technologies and available tools… … we make the benefits of the Web available to people who use simple mobile phones. ~~ ~~ ~~
- 94. Mobileusers VOICES User Content VOICES Platform & Toolbox Telecom Access Radio Program Output Web Access Mobile Training Lab Local ICT developers Speech Tools Local end-users Sustainable business models Mobile access PRODUCTS
- 95. Take home • Any ICT project which involves a user depends heavily on success of interface – Especially for ICT4D • User-centric design, participatory design, iterations • Cross-cultural HCI can teach us a lot • Action research, but agenda’s of ICT4D scientists need to be clear • Open mind – (low tech may be the best solution) • Dealing with low-literacy is complex, but very important
- 96. Questions?