Daniel Devatman Hromada

Berlin University of the Arts | UdK · Faculty of Design

This paper explores the "Rabbithole" instance of the Archaeoriddle Challenge, applying the NALANA method-a "naive, laic, and narrative" approach-to address the question: Was the relationship between poppy chewers (farmers) and rabbit skinners (hunter-gatherers) peaceful or hostile? The approach combined fairly common-sense logical reasoning with mo...

The Narrative Primer (NP) is an innovative, screen-less educational device designed to reform digitally assisted learning of children, especially those who are visually impaired or whose parents prefer audio-based educational media. Derived from post-smartphone Personal Primer [1], [2] artefact series, NP leverages advanced embedded-AI technologies...

We define an Extended Educational Environment (EEE or E3) as an immersive and interactive Extended Reality (XR) learning environment enriched with AI-coupled zones, artifacts and avatars which supports flexible configurations and multifunctional use to facilitate diverse educational experiences.

Wicked Solutions is a yearly compendium of scientific articles delivered by students of the course “Introduction to Optimization and Problem Solving in Coding and Making” which is a course offered by prof. Daniel D. Hromada to first semester Master students of the UdK / TU Study program Design & Computation.

In this article, we present three imperatives which help us to develop our AI-driven Personal Primer project in an ethically valid and sustainable manner. The deployment imperative states that one should not develop nor deploy AIED systems which one would be unwilling to use in learning process of one's own children. The alignment imperative requis...

During first part of "Berlin Symposium on Artificial Teacher-Student avatars"the first lecture, Daniel Hromada defined the notion of "Teacher Avatarizaion". Subsequently, during the hands-on exercises in cooperation with Christian Schmidts and Florian Porada, participants learned how to enrich a publicly available large language model (LLM) with an...

In our show & tell contribution, we present more closely seven properties of a digital educational artifact known as PersonalPrimer. These are: speech-based, embooked, voluminous, modular, circadian, SDG-compliant, edge-computing. Furthermore,we provide link to code repository as well as enumerate off-the-shelf components which can be combined toge...

Training, fine-tuning and other forms of supervised creation and optimization of speech recognition systems necessitate an existence of textual transcriptions describing the content of an associated audio file. In current speech signal datasets, text, speaker-metadata and audio signal itself are often stored in different files which increases compl...

The aim of the digital primer project is cognitive enrichment and fostering of acquisition of basic literacy and numeracy of 5-10 year old children. Here, we focus on Primer's ability to accurately process child speech which is fundamental to the acquisition of reading component of the Primer. We first note that automatic speech recognition (ASR) a...

The present study provides the first empiric evidence that the creation of human–machine peer learning (HMPL) couples can lead to an increase in the level of mastery of different competences in both humans and machines alike. The feasibility of the HMPL approach is demonstrated by means of Curriculum 1 whereby the human learner H gradually acquires...

In this short plaidoyer, we focus the attention of the reader on the fact that as of 2023, plethora of diverse platforms, apps and models exists which allow one to use AI-assisted systems in education of eco-and biological sciences. Nonetheless, these sciences are often left unnoticed by the AIED community and no curriculum seems to exist which wou...

The aim of the "Personal Primer", AI-driven project is to facilitate pupil's entry into the world of letters, numbers and codes. In this article, we present two imperatives which help us to design the Primer in an ethically valid and sustainable manner. The first imperative states that one should not develop nor deploy AIED systems which one would...

The aim of the digital primer project is cognitive enrichment and fostering of acquisition of basic literacy and numeracy of 5 – 10 year old children. Here, we focus on Primer's ability to accurately process child speech which is fundamental to the acquisition of reading component of the Primer. We first note that automatic speech recognition (ASR)...

Once upon a time, an academic paper had been written, expressing certain formal features isomorph to those of a fairy-tale. Essentially a discussion between a human author and a mainstream conservative ChatGPT-I model addressing the cultivation of notions of morality , bravery and justice in primary-school pupils: this half-narrative, half-dialogue...

The ultimate goal of the project AAA&O is to initiate transformation of a vaguely defined concept of “computational dramatics” into a young discipline combining the rigor of computer science with creativity and aesthetic aspirations of dramatic and performing arts. In concrete terms, this Goal is to be attained by staging of the semi-generative p...

Process of human learning has many features in common with the process of machine learning. This allows for creation of human-AI tandems or smaller groups where all members of the tandem or a group learn and develop. Consistently with Vygotskyan and Piagetian theories of learning and role which peers and intersubjective relations play in such theor...

This file is part B2 of project proposal "Digital Habits and Acquisition of Reading" unsuccessfully submitted to European Research Council's 2022 Starting Grant. DHAR's main objective is to assess strength and intensity of smartphone-induced habits among elementary school pupils and potentially falsify the zero hypothesis H0: "Strength of pupil's d...

Process of human learning has many features in common with the process of machine learning. This allows for creation of human-AI tandems or smaller groups where all members of the tandem or a group learn and develop. Consistenly with Vygotskyan and Piagetian theories of learning and role which peers and intersubjective relations play in such theori...

Departing from the definition of techno-addiction in terms of technology-assisted behaviour with probable detrimental consequences, we propose following guidelines could direct the design of harm-reducing technologies: gradual use-constraining, circadianity, offline preferentiality, environmental referentiality and monotasking. These guidelines can...

This study compares performance of different speech command classification systems which can be executed on an Raspberry Pi Zero ARMv6 architecture. Three systems are evaluated: first one, TREELITE_MFCC, is based on Treelite system cross-compiles a MFCC-classifying Random Forest into a highly optimized shared library; second, TFLITE_MFCC, performs...

teacher.js is a web-based digital communication and teaching tool. Developed as part of the teacher.solar project (whose goal is to take teaching outdoors, using entirely solar-powered devices) one of the main requirements was to keep both the bandwidth and power consumption low. For that reason teacher.js does not make use of video streaming (whic...

Departing from the postulate that Voynich Manuscript is not a hoax but rather encodes authentic contents, our article presents an evolutionary algorithm which aims to find the most optimal mapping between voynichian glyphs and candidate phonemic values. Core component of the decoding algorithm is a process of maximization of a fitness function whic...

Wie kann künstliche Intelligenz Kinder aktiv beim Lernprozess unterstützen?

Digital technologies permeate our every day life in an uncontrolable and accelerating manner. As of 2020, techno-addiction is already a global fact which concerns all generations, with youngest generation predicted to pay the highest cognitive price. It is impossible to stop this development but principles can be established and applied which are t...

Despite being an object of multi-billion public policies and private sector initiatives , the term "digital education" seems to lack a clear, unambigous, lexicon-ready definition. A closer analysis reveals that in common parlance, the term is used to denote phenomena related to overlapping but distinct topics like "technology", "media" or "informat...

In diesem Workshop-Beitrag beschreiben wir vier physische Eigenschaften eines digitalen Bildungsinstruments für Grundschüler, das als digitale Fibel bezeichnet wird. Diese sind: buchähnlich, voluminös, einzigartig und modular. Darüber hinaus zählen wir handelsübliche Bauteile auf, aus deren Kombination die digitale Fibel-Prototypen der ers-ten Gene...

Non-moderate smartphone usage may induce diverse pathological states and behaviors which may potentially result in an array of syndromes and illnesses. Digital devices built for education rather than consumption and entertainment should not neglect intricacies of human physiology, ergonomy and cognition. For this reason, we present first four prope...

Original research proposal for the Personal Primer project.

This document describes fundamental concepts and principles governing the transformation of an opaque content-publishing community economy into a fully transparent publicly available ledger. Inspired by existing public ledger technologies and cryptocurrencies like Ripple, our Kyberia’s Open Ledger (KOL) proposal aims to use minimalist variant of a...

This article presents method and results of multiple analyses of the biggest publicly available corpus of language acquisition data : Child Language Data Exchange System. The methodological aim of this article is to present a means how science can be done in a highly positivist, empiric and reproducible manner consistent with the precepts of the "O...

We introduce a novel method for transformation of texts into short binary vectors which can be subsequently compared by means of Hamming distance measurement. Similary to other semantic hashing approaches, the objective is to perform radical dimensionality reduction by putting texts with similar meaning into same or similar buckets while putting th...

Being based on English language, existing systems of part-of-speech induction prioritize the contextual and distributional features “external” to the word and attribute somewhat secondary importance to features derived from word’s “internal” morphologic and orthotactic regularities. Here we present some preliminary empirical results supporting the...

We present two axiomatic and three conjectural conditions which a model inducing natural language categories should dispose of, if ever it aims to be considered as "cognitively plausible". 1st axiomatic condition is that the model should involve a bootstrapping component. 2nd axiomatic condition is that it should be data-driven. 1st conjectural con...

Individual specificity and autonomy of a morally reasoning system is principally attained by means of a constructivist inductive process. Input into such process are moral dilemmata, or their story-like representations; its output are general patterns allowing to classify as moral or immoral even dilemmas which were not represented in the initial '...

Parallel democracy model is a variant of traditional participative democracy approach aiming to provide a new method of convergence toward quasi-optimal solutions of diverse perennial political and social challenges. Within the framework of the model, such challenges are operationalized into variables for which the authority can associate possible...

A language-independent method of figure-of-speech extraction is proposed in order to reinforce rhetoric-oriented considerations in natural language processing studies. The method is based upon a translation of a canonical form of repetition-based figures of speech into the language of PERL-compatible regular expressions. Anadiplosis, anaphora, anti...

Five different OpenCV-compatible XML haar cascades of zygomatic smile detectors as well as five SMILEsamples from which these detectors were derived had been trained and are presented hereby as a new open source SMILEsmileD package. Samples have been extended in an incremental learning fashion, exploiting previously trained detector in order to add...

The aim of our study was to show that distributions of hyperlinks within wikipedia corpora implicitly contain information about cultural preferences of its authors. We have transformed wikipedia corpora written in 27 different languages into graph structures whose vertices correspond to wikipedia articles and edges to hyperlinks between these artic...

The experiment consisted of three tasks: during first a spontaneous motor tempo (SMT) was measured; during the second a child had to synchronise to stimuli with 600ms interstimuli interval (ISI); third task was a continuation/induction task-after being attuned to a 600ms ISI, a child was instructed to continue tapping the same rhythm (IRI) even aft...

RÉSUMÉ Cet article présent le système hybride et multi-modulaire d'extraction des mots-clés à partir de corpus des articles scientifiques. Il s'agit d'un système multi-modulaire car intègre en soi les traitements 1) morphosyntaxiques (lemmatization et chunking) 2) sémantiques (Reflective Random Indexing) ainsi que 3) pragmatiques (modélisés par les...