A critical exploration of my digital footprint to unearth hidden aspects of self in online spaces.
A website that explores my digital identity through data mining and self-reflection.
UP.LOADED serves as a project to unearth my personal digital identity through a suite of experiments conducted using my own personal digital data from Facebook over 10 years. Being a research-based critical design project, these experiments are framed under an ethnography*, self-reflective lens.
Themes such as changes in personality over time, relationships and identity are explored and reflected upon—these experiments act as stimuli, which are paired with a reflective response in the form of code poetry, a form of creative poetry written under the rules of code/programming language that can still be understood by people not necessarily specialising in code.
Themes such as changes in personality over time, relationships and identity are explored and reflected upon—these experiments act as stimuli, which are paired with a reflective response in the form of code poetry, a form of creative poetry written under the rules of code/programming language that can still be understood by people not necessarily specialising in code.
*ETHNOGRAPHY:
“A branch of anthropology... the study of the beliefs, social interactions, and behaviours of small societies, involving participation and observation over a period of time, and the interpretation of the data collected.” Loshini Naidoo — An Ethnography of Global Landscapes and Corridors (2012).
“A branch of anthropology... the study of the beliefs, social interactions, and behaviours of small societies, involving participation and observation over a period of time, and the interpretation of the data collected.” Loshini Naidoo — An Ethnography of Global Landscapes and Corridors (2012).
A world where digital and physical start to blend.
We live in a world where digital identities play an increasingly crucial role in our lives and how our personalities are shaped as our physical and digital lives begin to blur. It has altered the way we approach interaction and connections in relationships and the way we work with and share information.
Additionally, our actions online exist in the form of indestructible data, and thus creates an ‘online archive’ of our actions, relationships and sense of identity. What can we learn about ourselves if we investigate this past? This idea of a perpetual digital space and morphing of identities will only become more significant as time goes on—so, it’s time to talk and find out about our data online.
This project acts as provocations to explore these questions by positioning myself as a case study and sharing my personal learnings and reflections—Our digital information is only a few clicks away, and UP.LOADED endeavours to spark that urge to take action and reflect on our digital information to discover new meanings and sense of self.
Additionally, our actions online exist in the form of indestructible data, and thus creates an ‘online archive’ of our actions, relationships and sense of identity. What can we learn about ourselves if we investigate this past? This idea of a perpetual digital space and morphing of identities will only become more significant as time goes on—so, it’s time to talk and find out about our data online.
This project acts as provocations to explore these questions by positioning myself as a case study and sharing my personal learnings and reflections—Our digital information is only a few clicks away, and UP.LOADED endeavours to spark that urge to take action and reflect on our digital information to discover new meanings and sense of self.
UP.LOADED - Carmen Yeh 2021
This project was produced by Carmen Yeh as part of the Visual Communication Honours degree at the University of Technology Sydney. Special thanks to Monica Monin and Zoë Sadokierski for assisting in initial experimentation and general guidance, and Alexandra Chalmers Braithwaite for significantly contributing with project development, direction and guidance. Credits to Maggie So for assisting with web development.
Framing the research method.
Framed under an auto-ethnographic* research approach, an array of methods were utilised, allowing me to explore rising technology culture and human digital experiences effectively through my own downloadable data and reflections. Placing myself as a case study enables me to reach a wider audience in a more relatable and personal manner, creating a more significant impact.
*AUTOETHNOGRAPHY:
A method of research where the writer draws from personal, anecdotal experiences through self-reflection to explore social and cultural meanings.
A method of research where the writer draws from personal, anecdotal experiences through self-reflection to explore social and cultural meanings.
Determining the data I am working with.
The source of all my data is from Facebook—A staggering 19 GB worth of information accumulated over a decade was downloaded. This information can be easily downloaded under the format of .JSON* files, allowing for easy data mining and sorting as it can be sorted under certain categories (timestamp, person, content, etc.). With this information in hand, I conducted a suite of experiments using a variety of processes. Shown below is the structure of the downloaded data.
*.JSON (JavaScript Object Notation)
A text-based file type that is able to contain human-readable data that can also be effectively stored and transmitted by machines.
A text-based file type that is able to contain human-readable data that can also be effectively stored and transmitted by machines.
Both primary and secondary processes were involved.
To analyse and dissect this information, data collating, mining and visualisation were the primary processes utilised to explore previously concealed trends, patterns, or any unusual/new findings.
These processes were conducted through the use of code, specifically Python* as it is the most accessible language for data analysis. Libraries such as Textblob was also utilised which assisted in conducting sentiment analysis*.
In the context of this project, data mining includes processes such as:
With these primary processes used for the actual experimentation acting as stimuli, secondary methods were utilised as a response to these findings.
Such methods include user journey mapping, critical reflection through journaling, annotating, diagramming and user-testing to explore different methods of reflection and contemplation.
These processes were conducted through the use of code, specifically Python* as it is the most accessible language for data analysis. Libraries such as Textblob was also utilised which assisted in conducting sentiment analysis*.
In the context of this project, data mining includes processes such as:
- Determining the usage/frequency of certain words used
- Conducting sentiment analysis on words—whether certain messages are positive, neutral or negative in terms of emotions
- Visual analysis & mapping of screenshots/photos to uncover patterns
- Message activity frequency based on time of day
With these primary processes used for the actual experimentation acting as stimuli, secondary methods were utilised as a response to these findings.
Such methods include user journey mapping, critical reflection through journaling, annotating, diagramming and user-testing to explore different methods of reflection and contemplation.
*PYTHON:
A non-specific programming language that is typically used for website building, software development and data analysis.
*SENTIMENT ANALYSIS:
A method of analysis (often done with code) to identify the mood and emotive value behind a piece of writing/text.
A non-specific programming language that is typically used for website building, software development and data analysis.
*SENTIMENT ANALYSIS:
A method of analysis (often done with code) to identify the mood and emotive value behind a piece of writing/text.
Presenting the data.
As the main concept in all of these experiments involve change over time, each experiment is presented in chronological order from past to present over the 10 years of data.
Each experiment is presented across the web page with a code poem overlayed on top to create a multilayered user-experience where the viewer can read both the stimuli and response simultaneously. Through this method of recontextualising by presenting 10 years of data in one sitting, a new perspective of time is formed.
Each experiment is presented across the web page with a code poem overlayed on top to create a multilayered user-experience where the viewer can read both the stimuli and response simultaneously. Through this method of recontextualising by presenting 10 years of data in one sitting, a new perspective of time is formed.
UP.LOADED - Carmen Yeh 2021
A number of research anchors* has sparked different possible ways data can be experimented with and presented unconventionally.
Memory Metadata - Livia Foldes (2019)
Memory Metadata is a project where Foldes reflects on points of her life through the analysis of pixels in personal photos. She investigates the idea of looking into her life from a digital standpoint and branches out in an attempt to spark conversations surrounding ethical concerns regarding AI.
This anchor was a significant influence on the process and idea behind LOVE.LANGUAGE, where photo analysis was conducted to explore how love can be communicated digitally.
Memory Metadata is a project where Foldes reflects on points of her life through the analysis of pixels in personal photos. She investigates the idea of looking into her life from a digital standpoint and branches out in an attempt to spark conversations surrounding ethical concerns regarding AI.
This anchor was a significant influence on the process and idea behind LOVE.LANGUAGE, where photo analysis was conducted to explore how love can be communicated digitally.
*ANCHOR:
An existing project or piece of writing (precedent) which has contributed to the project direction and development.
Term coined by Zoë Sadokierski — Developing critical documentation practices for design researchers (2020).
An existing project or piece of writing (precedent) which has contributed to the project direction and development.
Term coined by Zoë Sadokierski — Developing critical documentation practices for design researchers (2020).
Tech and me: an autoethnographic account of digital literacy as an identity performance - Daniel Clark (2020)
In this paper, Clark investigates changes in his identity by conducting autoethnographic research on blog posts he has written over 8 years.
This article explored the concept of change over time, which has become a significant driving force in my project. His autoethnographic methodology involving reviewing his past posts on social media also served as a significant influence towards the processes used in my experiments, specifically, HAPPY.BIRTHDAY.
In this paper, Clark investigates changes in his identity by conducting autoethnographic research on blog posts he has written over 8 years.
This article explored the concept of change over time, which has become a significant driving force in my project. His autoethnographic methodology involving reviewing his past posts on social media also served as a significant influence towards the processes used in my experiments, specifically, HAPPY.BIRTHDAY.
Code {poems} - Ishac Bertran (2012)
Code {poems} is a book which involves collaboration with artists and software engineers to explore code written in the form of a narrative by combining it with human language, where its message is up to the readers’ interpretation.
This publication explores how the combination of machine and human language can bring about new reading experiences. In my project, I drew from this concept to bring about a unique and compelling way to convey my personal reflections. At the same time, writing under the influence of code language allows for a non-invasive way
to express my thoughts and bring it out to the world to tell.
Below is an overview map showcasing the influence from research anchors for each experiment.
Code {poems} is a book which involves collaboration with artists and software engineers to explore code written in the form of a narrative by combining it with human language, where its message is up to the readers’ interpretation.
This publication explores how the combination of machine and human language can bring about new reading experiences. In my project, I drew from this concept to bring about a unique and compelling way to convey my personal reflections. At the same time, writing under the influence of code language allows for a non-invasive way
to express my thoughts and bring it out to the world to tell.
Below is an overview map showcasing the influence from research anchors for each experiment.
UP.LOADED - Carmen Yeh 2021
LOVE.LANGUAGE
Investigating how love is communicated from a digital standpoint.
LOVE.LANGUAGE is an experiment which involves photo and screenshot analysis by visually arranging all media that I have interacted with online (both sent or received) over time into a grid. Parameters were set when arranging these photos such as time sent/received and the type of media.
The final showcase of this particular experiment focused on the idea of a digital love language—how this can be communicated through sharing photos. By arranging the photos and identifying each food photo, I was able to concurrently visualise how much I communicate feelings with food in the digital space.
To visually present these findings, each food related media was highlighted by replacing them with white squares. This is made interactive where each square reveals what that particular photo entailed when hovered over.
LOVE.LANGUAGE pushes the boundaries of what a love language can be, and enabled me to make sense of how I express affection and the nuanced actions I take to connect with others online.
The final showcase of this particular experiment focused on the idea of a digital love language—how this can be communicated through sharing photos. By arranging the photos and identifying each food photo, I was able to concurrently visualise how much I communicate feelings with food in the digital space.
To visually present these findings, each food related media was highlighted by replacing them with white squares. This is made interactive where each square reveals what that particular photo entailed when hovered over.
LOVE.LANGUAGE pushes the boundaries of what a love language can be, and enabled me to make sense of how I express affection and the nuanced actions I take to connect with others online.
FREQUENT.WORDS
An exploration of how choices of unique words change over time.
Methods involved in FREQUENT.WORDS include data mining and word analysis where code is utilised to find out the most used words when I message others online. These data minings were also conducted several times to explore how these findings would change under different time periods.
This is visually presented with a list of words along with the number of times they appear in my online conversations in chronological order.
An individual’s choice of words serve as indications of their character. By analysing my use of words over time, I was able to witness how I have developed as a person over time. These changes included a rising trend in word length, and words with strong emotive connotations.
This is visually presented with a list of words along with the number of times they appear in my online conversations in chronological order.
An individual’s choice of words serve as indications of their character. By analysing my use of words over time, I was able to witness how I have developed as a person over time. These changes included a rising trend in word length, and words with strong emotive connotations.
HAPPY.BIRTHDAY
Unearthing the intention behind each Facebook post.
Applying a similar methodology to Daniel Clark’s Tech and me, HAPPY.BIRTHDAY investigates the authenticity behind Facebook posts, and the changes in content that is posted over time.
When investigating my Facebook posts, a staggering number of birthday posts had appeared. The method of close reading* was used to highlight the amount of times I had posted a ‘generic’ birthday post that could’ve been addressed to anyone. This analysis involved laying all of my posts into one space, and connecting each birthday post with a line to generate a visual pattern.
This is visually portrayed by having each white dot signifying a birthday post, with lines interconnecting each one. This puts into perspective the rise in generic posts over time, raising conversations about interaction with others online and the gradual shift in the use of public social media.
![]()
![]()
When investigating my Facebook posts, a staggering number of birthday posts had appeared. The method of close reading* was used to highlight the amount of times I had posted a ‘generic’ birthday post that could’ve been addressed to anyone. This analysis involved laying all of my posts into one space, and connecting each birthday post with a line to generate a visual pattern.
This is visually portrayed by having each white dot signifying a birthday post, with lines interconnecting each one. This puts into perspective the rise in generic posts over time, raising conversations about interaction with others online and the gradual shift in the use of public social media.
*CLOSE READING:
“Thoughtful, critical analysis of a text that focuses on significant details or patterns in order to develop a deep, precise understanding of the text’s form, craft, meanings, etc.” Beth Burke — A Close Look at Close Reading (n.d.).
“Thoughtful, critical analysis of a text that focuses on significant details or patterns in order to develop a deep, precise understanding of the text’s form, craft, meanings, etc.” Beth Burke — A Close Look at Close Reading (n.d.).
TRACK.RECORD
Unveiling the emotive states of each message sent online.
Using a similar process as FREQUENT.WORDS, TRACK.RECORD measures the emotive states of each message I’ve sent and received across three close friends. Through code, a sentiment analysis* was conducted, where a value between -1 to 1 (negative to positive emotion) is determined for each message.
This data was further experimented with through data visualisation, where one of the iterations was chosen to portray the information.
TRACK.RECORD was an experiment which prompted self-reflection on the states of my personal friendships, and how I portray myself towards different people online—I particularly found the stark contrast in polarities between two people very jarring, where one relationship had very consistent values, and and the other having many extreme values.
![]()
![]()
![]()
This data was further experimented with through data visualisation, where one of the iterations was chosen to portray the information.
TRACK.RECORD was an experiment which prompted self-reflection on the states of my personal friendships, and how I portray myself towards different people online—I particularly found the stark contrast in polarities between two people very jarring, where one relationship had very consistent values, and and the other having many extreme values.
Things that didn’t make the cut but are worth the mention.
UP.LOADED - Carmen Yeh 2021
UP.LOADED - Carmen Yeh 2021
Utilising reflection as a method.
As a critical part of the process, my personal reactions, findings and reflections are recorded through methods including reflective writing, keeping experimental logs and journaling.
A response to the experiments written under code language.
In order to convey these personal findings and reflections for each experiment, they are presented in the form of code poetry as a response to each stimuli.
Multiple iterations were done with changes based on feedback by presenting drafts to people who are not familiar with code to ensure the poems are accessible and understood by a wide, general audience.
The combination of human and code language ultimately leads to an alternative reading experience for viewers without straying too far into unfamiliar territory. With an slight barrier naturally forming due to the nature of code syntax and language patterns, I was able to convey my personal thoughts without the risk of over-exposure and maintaining a degree of privacy.
Multiple iterations were done with changes based on feedback by presenting drafts to people who are not familiar with code to ensure the poems are accessible and understood by a wide, general audience.
The combination of human and code language ultimately leads to an alternative reading experience for viewers without straying too far into unfamiliar territory. With an slight barrier naturally forming due to the nature of code syntax and language patterns, I was able to convey my personal thoughts without the risk of over-exposure and maintaining a degree of privacy.
UP.LOADED - Carmen Yeh 2021
UP.LOADED - Carmen Yeh 2021