AI Powered Self Reflection
Aydanur Akkut
Humans are born into an unending quest for self-discovery. Through algorithms and big data, everything, from our preferences to our habits, and even our most hidden fears, can be transformed into numbers and models. However, obtaining meaningful findings requires large and cleaned datasets.
This study aims to explore whether new technologies, through the use of artificial intelligence tools and coding, can hold a mirror to a person’s life and emotions during the process of self-discovery. In this context, approximately 40 video diaries recorded at various times and intervals over four years were analyzed. Through this analysis, the individual’s personal development process and emotional shifts were examined. Three different types of data were utilized in this study, which are text, audio, and image.
The data has been examined in two different forms: readable data and data art.
Readable Data
Texts, speech transcripts, and facial expressions found in the individual’s diaries were analyzed using artificial intelligence tools and Python, then converted into meaningful outputs. This data was examined through measurable characteristics such as emotional shifts, recurring themes, and mental orientations. The objective is to reveal traces of the person’s inner world in a more objective and systematic manner.
You can navigate between tabs and examine the charts interactively.
By reviewing the charts, you can observe which emotions were felt in which years based on the analyzed diaries.
The artist aims to encourage the viewer to reflect on the following questions: “In the journey of self-discovery, if a mirror were held up to a person using randomly recorded personal data, what would be seen? Can artificial intelligence and code truly know a person well enough to reflect them back to themselves? How can new technologies assist an individual in self-recognition and in reviewing their processes? Is it beneficial or unsettling for algorithms to possess personalized data?”
The 50 most used words by year (excluding stop words). (The larger the size of the word, the higher the frequency of its usage.)
2021

2023

2022

2024

For each video, facial expressions were labeled by capturing a frame every 60 frames throughout the duration. A personalized dataset was created by associating facial expressions with 9 different emotions (happy, angry, sad, surprised, excited, disgusted, neutral, scared, anxious). These photos were assembled using a collage technique.

Data Art
The aim was to approach data not only through its measurable aspects but also through its emotional and aesthetic dimensions. Within this scope, facial expressions derived from the videos and imagery found in the texts were reinterpreted into artistic forms. Visual compositions were created to allow emotions to establish an intuitive bond with the viewer, transcending mere numbers. Thus, data was transformed into creative narratives that express personal experience through storytelling.
You can navigate between tabs and examine the charts interactively.
An emotion map was created by positioning labels on a 2D plane using a mathematical formula based on the targeted (intended to emerge) visual. The raw form of the data was shaped using data visualization tools and basic trigonometric formulas.
Upon this technical infrastructure, an artistic language was developed by adding the felt aspects of emotions to their measurable ones. Data was reinterpreted through color, form, and movement; emotional fluctuations over the years were transformed into intuitively readable visual compositions.
Thus, the aim was to transform numerical data into a personal narrative and experience, establishing a more direct bond with the viewer.
You can navigate between tabs and examine the charts interactively.
The artist tried to create alternatives for self-discovery and expression using modern technologies (coding, artificial intelligence, etc.). She aims to encourage the viewer to explore their own emotions and express them in different forms.

Aydanur AKKUT (she/they)
Experimental Artist
Born in 2000 in İzmir. She is an engineer and data analyst. Her work encompasses subjects such as data analysis, the digitization of production processes, and the adaptation of artificial intelligence into creative processes.
She is closely interested in intersectional activism. She explores themes of collective memory, belonging, and empathy, departing from individual experience. She is the founder and curator of the Identity 5.0 Exhibition. She aims to design multidisciplinary spaces that encourage the participation of individuals and communities in creative processes.
