User:Alexander Roidl/bookscanning: Difference between revisions

From XPUB & Lens-Based wiki
No edit summary
No edit summary
 
(11 intermediate revisions by the same user not shown)
Line 27: Line 27:
* generating new data  
* generating new data  
(endless data generation)
(endless data generation)
* what data is selected?
* how to find sth. in that data (analysis)?
* FILTERS


=bubbles=
=bubbles=
Line 56: Line 59:
* quantitative method (counting women vs. counting men)
* quantitative method (counting women vs. counting men)


==Sources==
 
 
 
=Feedback=
 
* scanned material to train an algorithm > a chatbot trained on the books that are scanned > ref to Google Books (and to failed twitterbot)
** always limited resources as input
* What happens if you train this chat bot on a book, not on social media material
 
Connecting to the reader
* What is lost, what is gained by scanning
* forensic materiality to more performative materiality
 
* "Not enough data" as a mantra to promise that the algorithm gets better
* The scanner as a mediator to use other types of data
 
> fembot
https://www.nytimes.com/2018/02/04/arts/fembot-poppy-lil-miquela-kylie-jenner.html
 
http://soulellis.com/ book on chatbots by artists and designers
 
Enron company, the Enron dataset https://www.cs.cmu.edu/~enron/
datasets that are used often also shape the language, a standardized way to assess 'natural' language
leaked data used for machine learning
Sam Lavigne and Tega Brain: The Good Life
https://www.newmuseum.org/calendar/view/1258/the-making-of-natural-language (there is also a video recording of the event)
 
http://publicdomainday.constantvzw.org/  the death of the author
The Bot Opera http://video.constantvzw.org/PublicDomainDay2015/NEWNOVA.webm
2015 edition: http://publicdomainday.constantvzw.org/#1943
 
Train chatbot on dataset
-> extremist references
-> introduction on
 
=the database / data-usage=
 
how does the data influence the algorithm?
 
where data for algorithms is coming from
-> books / internet
-> training AI
 
how is the data being selected?
 
 
-> you can not see the data in the background
 
-> you do not know what Google uses its data for
 
 
==Feedback Steve==
 
* Protocol by Alexander Gallaway (how interface is organized)
* Software takes command (serve interests of several groups)
* Sorting Things Out
* Literacy and its consequences
 
==Related Texts==
Sorted by importance
 
* Lev Manovich – The database
* http://adanewmedia.org/blog/2015/11/01/issue8-masters/
* http://adanewmedia.org/blog/2016/05/01/issue9-hoffmann-and-bloom/
* webs of feminist knowledge
* https://www.wired.com/2010/02/ff_google_algorithm/
 
 
 
 
==Femenist methodologies==
 
=Sources=


The Mass Digitization of Books:
The Mass Digitization of Books:
Line 74: Line 149:
Article on Roger Chartier and the past & Future of books
Article on Roger Chartier and the past & Future of books
http://www.booksandideas.net/The-Book-Its-Past-Its-Future.html
http://www.booksandideas.net/The-Book-Its-Past-Its-Future.html
Data selection of AI bots
https://www.newmuseum.org/exhibitions/view/sam-lavigne-and-tega-brain-the-good-life
https://www.cs.cmu.edu/~enron/
Fembot
https://www.nytimes.com/2018/02/04/arts/fembot-poppy-lil-miquela-kylie-jenner.html

Latest revision as of 14:25, 7 February 2018

IMG 6602.jpg

transformative processes

  • transforming knowledge
  • creating new file
  • creating new »content«
  • new possibilities of processing (OCR, analysis…)
  • revealing information layers (printing technique)
  • turning an analog medium into a digital file
    • change of object, color, pages, reading flow
    • Kantian model of the book (object, that’s sellable & knowledge)
    • Chartier
  • display is changing
  • reproducing the reproduced (Walter Benjamin -> next level)

scanning as a tool

  • creating graphics / visuals out of scanned pages

-> draw an image from a scanned book pages -> AI? drawing images / abstract / concrete / emotion

  • missuse -> new unexpected results

big data

  • generating new data

(endless data generation)

  • what data is selected?
  • how to find sth. in that data (analysis)?
  • FILTERS

bubbles

  • we live within a given framework
  • selection of texts leads to an opinion
  • filter
  • idea: chatbot trained only from scanned book pages
  • sherry turkle?

light-sensitive color

  • print with light-sensitive color

-> see nothing / image when scanned or -> see the text when scanned, but it disappears after scanning

More

Access to…

  • digitale books
  • analog books
  • who has access / who hasn't and why?
  • how easy is access (com. digital / analog)

Adding/Deleting/Altering information of existing books

  • (piracy - connected to anonymous )

Feminist Methodologies

How digitization is changing the view on sexism/feminism

  • books become searchable and more connected
  • quantitative method (counting women vs. counting men)



Feedback

  • scanned material to train an algorithm > a chatbot trained on the books that are scanned > ref to Google Books (and to failed twitterbot)
    • always limited resources as input
  • What happens if you train this chat bot on a book, not on social media material

Connecting to the reader

  • What is lost, what is gained by scanning
  • forensic materiality to more performative materiality
  • "Not enough data" as a mantra to promise that the algorithm gets better
  • The scanner as a mediator to use other types of data

> fembot https://www.nytimes.com/2018/02/04/arts/fembot-poppy-lil-miquela-kylie-jenner.html

http://soulellis.com/ book on chatbots by artists and designers

Enron company, the Enron dataset https://www.cs.cmu.edu/~enron/ datasets that are used often also shape the language, a standardized way to assess 'natural' language leaked data used for machine learning Sam Lavigne and Tega Brain: The Good Life https://www.newmuseum.org/calendar/view/1258/the-making-of-natural-language (there is also a video recording of the event)

http://publicdomainday.constantvzw.org/ the death of the author The Bot Opera http://video.constantvzw.org/PublicDomainDay2015/NEWNOVA.webm 2015 edition: http://publicdomainday.constantvzw.org/#1943

Train chatbot on dataset -> extremist references -> introduction on

the database / data-usage

how does the data influence the algorithm?

where data for algorithms is coming from -> books / internet -> training AI

how is the data being selected?


-> you can not see the data in the background

-> you do not know what Google uses its data for


Feedback Steve

  • Protocol by Alexander Gallaway (how interface is organized)
  • Software takes command (serve interests of several groups)
  • Sorting Things Out
  • Literacy and its consequences

Related Texts

Sorted by importance



Femenist methodologies

Sources

The Mass Digitization of Books: http://www.kcoyle.net/jal-32-6.html

https://en.wikipedia.org/wiki/Roger_Chartier

Discussion on Googles Book Scanning http://www.firstmonday.org/article/view/2101/2037

NYT article on Google’s bookscanning & bookscanning in general http://www.nytimes.com/2006/05/14/magazine/14publishing.html

(Video) What is a book by Roger Chartier http://mediasitemob1.mediagroup.ubc.ca/Mediasite/Play/b7ff39a973534d4c9583fcd935ee2c581d

Article on Roger Chartier and the past & Future of books http://www.booksandideas.net/The-Book-Its-Past-Its-Future.html

Data selection of AI bots https://www.newmuseum.org/exhibitions/view/sam-lavigne-and-tega-brain-the-good-life

https://www.cs.cmu.edu/~enron/

Fembot https://www.nytimes.com/2018/02/04/arts/fembot-poppy-lil-miquela-kylie-jenner.html