User:Manetta/graduation-proposals/proposal-0.1

graduation proposal +0.1.1

title: "i could have written that"

alternatives:

• turning words into numbers
• machine-human-machine

Introduction

For in those realms machines are made to behave in wondrous ways, often sufficient to dazzle even the most experienced observer. But once a particular program is unmasked, once its inner workings are explained in language sufficiently plain to induice understanding, its magic crumbles away; it stands revealed as a mere collection of procedures, each quite comprehensible. The observer says to himself "I could have written that". With that thought he moves the program in question from the shelf marked "intelligent" to that reserved for curios, fit to be discussed only with people less enlightened that he. (Joseph Weizenbaum, 1966)

what do you want to do?

• setting up a publishing platform / magazine to reveal inner workings of technologies that systemize language.

Relation to previous practice

what are you doing?

In the last year, i've been looking at different tools that contain linguistic systems. From speech-to-text software to text-mining tools, they all systemize language in various ways in order to understand natural language—as human language is called in computer science. These tools fall under the term 'Natural Language Processing' (NLP), which is a field of computer science that is closely related to Artificial Intelligence (AI).

As a continutation of that i took part at the Relearn summerschool in Brussels last August, to propose a working track in collaboration with Femke Snelting on the subject of 'training common sense'. With a group of people we have been trying to deconstruct the truth-construction in algorithmic cultures, by looking at data mining processes, deconstructing the mathematical models that are used, finding moments where semantics are mixed with mathematic models, and understanding which cultural context is created around this field.

Another entrance to understanding what happens in algorithmic practises such as machine learning, is by looking at training sets that are used to train software that is able to recognize patterns that are trained by an algorithm. These training set could contain a large set of images, text, 3d models, or video's. By looking at such databases, and more specifically at the choises that have been made in terms of structure and hierarchy, certain steps of the construction a certain 'truth' are revealed.

There are a few datasets in the academic world that seem to be basic resources to built these trainin sets upon. In the field they are called 'knowledge bases'. They live on a more abstract level then the training sets do, as they try to create a 'knowlegde system' that could function as a universal structure. Examples are WordNet (a lexical dataset), ConceptNet, and OpenCyc (an ontology dataset).

Relation to a larger context

Thesis intention

Practical steps

how?

• following the approach of 'i will tell you everything' (my truth is a constructed truth), which took the structure of a machine learning training set (called the SUN dataset), and applied this to a set of objects that formed an exhibition together. The training set was (also literally) a voice over of this exhibition, that framed the objects by speaking from the choices that have been made to construct the SUN dataset.

• writing/collecting from a technology point of departure, to relate the topic of 'systemization/abstraction/modelling/'

references: 
- Matthew Fuller, powerpoint
- Constant, pipelines
- Steve Rushton, feeback
- Angie Keefer, Octopus

References

datasets

* WordNet (Princeton)
* ConceptNet 5 (MIT Media)
* OpenCyc

people

algorithmic culture

Luciana Parisi
Matteo Pasquinelli
Antoinette Roivoy
Seda Gurses 

other

Matthew Fuller

reading list

notes and related projects

BAK lecture: Matthew Fuller, on the discourse of the powerpoint (Jun. 2015) - annotations

project: Wordnet

project: i will tell you everything (my truth is a constructed truth)

project: serving simulations

---

graduation proposal +0.1.2

title: #!PATTERN+

Introduction

what are you doing?

Last January (2015) an interdisciplinary arts-lab in Brussels called Constant organized a worksession called 'Cqrrelations' in which we reflected on the construction of machine learning models in the field of text-mining. Together we learned about a software package developed by the CLiPS research centre of the University of Antwerp, called Pattern. With a smaller group we trained an algorithm to tell us if a certain text could be called 'patternalistic' or not.

As a continutation of that i took part at the Relearn summerschool in Brussels last August, to propose a working track in collaboration with Femke Snelting on the subject of 'training common sense'. With a group of people we have been trying to deconstruct the truth-construction in algorithmic cultures, by looking at data mining processes, deconstructing the mathematical models that are used, finding moments where semantics are mixed with mathematic models, and understanding which cultural context is created around this field.

* understanding algorithmic processes,
* questioning where choises are made for the construction
* finding moments where semantics are mixed with math
* looking at the data-mining culture

what do you want to do?

• publishing about this truth construction process in algorithmic cultures, by taking Pattern as a case-study object

A way of speaking back to these algorithmic cultural fields would be by publishing a critical fork of the text-mining software package called Pattern. The fork will be called #!PATTERN+, which will be a new release of the original package developed by the CLiPS research group at the university of Antwerp.

Relation to previous practice

Relation to a larger context

Thesis intention

Practical steps

how?

The critical fork '#!PATTERN+' will contain annotations in the form of commands inside the code, files that reflect questions that arose, and alternative tutorials will be added to the original package. By asking the question how is algorithmic 'truth' constructed?, these #!PATTERN+ notes will be developed in one of the following 3 sub-fields:

• Knowledge Discovery in Data (KDD) steps - a technical approach, revealing the process from 'raw' data to the presentation of the results
• text mining case studies - a reflection on the research projects that gave been done/are going on, working with examples and demo's
• culture of data-mining - a more context-based approach, looking at the communication conventions that are present in the field of data-mining

References

people

Luciana Parisi Matteo Pasquinelli Antoinette Roivoy Seda Gurses

events

Cqrrelations (Jan. 2015)

BAK lecture: Algorithmic Culture (Jun. 2015) - annotations

software

* CLiPS Pattern, official website

reading list

via Femke: Seda [Gurses preparing a session on Machine Learning]

notes and related projects

Transmediale lecture: All Watched Over By Algorithms (Jan. 2015) - annotations

earlier notes (Jul. 2015)

notes taken during Relearn Aug. 2015, training common sense

#!PATTERN+.readme

the Annotator, Cqrrelations Jan. 2015, on etherpad

notes Cqrrelations day 1, day 2, day 3, day 4, day 5