Loading...
Title: Humanizing Technology, Vol. III

Humanizing Technology, Vol. III

Artificial Intelligence and the Humanities

by Silvia Lima Gonçalves Araújo (Volume editor) Micaela Aguiar (Volume editor) Dalila Durães (Volume editor)
©2026 Edited Collection XVIII, 234 Pages
Science, Society & Culture
Series: Digital Learning and the Future, Volume 5

Summary

This third volume of the Humanizing Technology collection explores the intersections between artificial intelligence (AI) and the humanities, addressing issues ranging from language processing, ethics, and data practices to digital heritage and computational creativity. By combining contributions from multiple disciplines, it offers diverse perspectives on how AI is reshaping research, creativity, and culture. Each chapter stands alone while also contributing to a broader reflection on the evolving role of AI in scholarly and artistic practices. This volume is an essential reference for researchers, educators, and practitioners interested in the transformative impact of AI on the humanities.

Table Of Contents

  • Cover
  • Title Page
  • Copyright Page
  • Contents
  • List of Figures
  • List of Tables
  • Acknowledgments
  • Preface
  • Part I AI and Language Processing
  • 1 Sympathy for the Devil? AI-run Proficiency Tests (Gabriel Martins)
  • 1. First Words
  • 2. Literacy in the Digital Era
  • 3. Theory of Affordances: Possibilities and Hazards Considered
  • 4. Study Methodology Outline
  • 5. Considerations on DET’s Technical Manual
  • 6. Closing Remarks
  • Acknowledgments
  • Author Note
  • References
  • 2 Uncovering Inconsistencies in AI Text Analysis: A Review of ChatGPT 4.0’s Stop Word Frequency Counting (João Pedro Oliveira)
  • 1. Introduction
  • 2. Literary Review
  • 3. Methodology
  • 3.1. Using ChatGPT 4.0 for word frequency counting
  • 4. Data Analysis and Results Discussion
  • 4.1. Irregularities evaluation
  • 4.2. Null cases in word frequency counting
  • 5. Conclusion
  • Author Note
  • References
  • 3 Comparing the Ability of Generative Language Models to Understand Medieval Languages (Gergely Pethő, Ragini Menon, Matthias Schöffel, and Sabine Tittel)
  • 1. Introduction
  • 2. Related Work
  • 3. Materials
  • 4. Method
  • 4.1. Language models used
  • 4.2. Corpus preprocessing
  • 4.3. Manual assessment of translation quality
  • 4.4. Sentence-level alignment between MT and reference
  • 4.5. Prompting
  • 4.6. Evaluation
  • 5. Results
  • 6. Conclusions
  • 7. Data Availability
  • Acknowledgments
  • Author Note
  • References
  • 4 An Intelligent Approach to Aligning LIBRAS and LGP Signs and Metatexts in Portuguese Using Generative AI (Alexia Duarte, Sérgio Manuel Serra da Cruz, and Silvia Lima Gonçalves Araújo)
  • 1. Introduction
  • 2. Related Works
  • 3. Ludwig Architecture
  • 4. The Ludwig Platform and Digital Humanities
  • 5. Experimental Results and Discussion
  • 6. Conclusion
  • Author Note
  • References
  • 5 The Implementation of Brazilian Academic Multiliteracy Corpus (BrAMCorp): Toward a Systematic Investigation of Event and Process Data (Inês Signorini and Rodrigo Esteves de Lima-Lopes)
  • 1. The Brazilian Academic Writing Corpus
  • 2. Background and Foundations
  • 3. Toward a Systematic Investigation of Event and Process Data
  • 3.1. Research shift
  • 3.2. Proposed actions for future development
  • 4. Concluding Remarks
  • Author Note
  • References
  • 6 Enigma: Development of an Integrated Platform for Natural Language Processing and Sentiment Analysis for the Portuguese Language (Lilian Carvalho Oliveira and Mônica Ferreira da Silva)
  • 1. Introduction
  • 2. Methodology
  • 2.1. Objective
  • 2.2. Development of the Sentiment Analysis Function
  • 2.3. Development of the Word Class Functions
  • 2.4. Development of the Text Readability Function (Coleman–Liau index)
  • 2.5. Development of the Lexical Diversity Function
  • 2.6. Development of Density and Lexical Sophistication Functions
  • 3. Results and Discussion
  • 4. Conclusions and Limitations
  • Author Note
  • References
  • 7 “Misinformation” Everywhere but Nowhere: The Use of Topic Modeling and Generative AI in Linguistics (Michelle Rufrano, Jean-Ezra Yeung, Katrin Herget, and Teresa Alegre)
  • 1. Introduction
  • 2. Methodology
  • 2.1. Misinformation terms
  • 2.2. Summary statistics
  • 2.3. Qualitative data analysis
  • 3. Results
  • 4. Conclusion
  • Author Note
  • References
  • Part II AI, Ethics, and Data Practices
  • 8 Identifying Gender Bias Across Urdu Language Pre-trained Models (Qaisar Aziz, Adeel Zafar, and Imad Ullah)
  • 1. Introduction
  • 2. Related Work
  • 2.1. RoBERTa and BERT models for Urdu language
  • 2.2. Urdu text dataset
  • 3. Proposed Methodology
  • 3.1. Dataset
  • 3.2. NLP models for Urdu text sentiment analysis
  • 4. Experimental Results
  • 4.1. Prediction of Urdu text sentences through RoBERTa
  • 4.2. Predication of Urdu text sentences through BERT
  • 4.3. Models predictions on sentimental sentences
  • 4.4. Sample of predicted sentences RoBERTa and BERT
  • 5. Conclusion and Future Work
  • Author Note
  • References
  • 9 Algorithmic Inequalities: A Comparative Methodology for Toxicity Assessment (Daniel Bonatto Seco and Leandro Guimarães Marques Alvim)
  • 1. Contextualization
  • 1.1. Fairness and bias
  • 1.2. Risk mitigation approaches
  • 1.3. Related studies
  • 2. A Comparative Methodology for Bias Identification
  • 3. Conclusion
  • Author Note
  • References
  • 10 Generative Artificial Intelligence and UNESCO’s Role: Insights Through Edgar Morin’s Complexity Theory (Aline Barbosa Xavier Muniz)
  • 1. Introduction
  • 1.1. UNESCO’s leadership in AI for higher education
  • 1.2. Contribution of complexity theory: A human-centered perspective on AI
  • 2. Conclusion
  • Author Note
  • References
  • 11 AI-based Chatbots as a Customer Communication Tool: Comparative Analysis between Germany and China (Anja Corduan-Claussen)
  • 1. Introduction
  • 2. Theoretical Background
  • 2.1. Fundamental concepts of chatbots
  • 3. Development of ChatGPT
  • 4. Development of ERNIE
  • 5. Cost Factors in the Implementation and Use of Chatbots
  • 5.1. Development costs
  • 5.2. Infrastructure costs
  • 5.3. Operating costs
  • 6. Methodology
  • 7. Utilization of Chatbots
  • 7.1. Application areas in Germany
  • 7.2. Application areas in China
  • 8. Acceptance of Chatbots
  • 8.1. Acceptance in German society
  • 8.2. Acceptance in Chinese society
  • 9. Comparison of Germany and China
  • 9.1. Ethical challenges in Germany
  • 9.2. Ethical challenges in China
  • 9.3. Cultural frameworks
  • 9.4. Infrastructure
  • 9.5. Practical considerations
  • 10. Logistics and Actual Application
  • 11. Limitation and Further Research
  • 12. Conclusion
  • Author Note
  • References
  • 12 Bridging Disciplines Among Digital Humanities, Cybercrime Forensics, and Artificial Intelligence (Emerson Duarte and Sergio Manuel Serra da Cruz)
  • 1. Introduction
  • 2. Cybercrimes Today
  • 3. Digital Humanities, Digital Forensics, and Artificial Intelligence
  • 4. Systematic Literature Review
  • 4.1. Stages of the systematic review
  • 4.1.1. Objectives and research question
  • 4.1.2. Database selection
  • 4.1.3. Search string definition
  • 4.1.4. Screening criteria
  • 4.1.5. Results found
  • 4.1.6. Quality criteria
  • 4.2. Results of the systematic search procedures.
  • 5. Analysis and Discussion of the Selected Studies
  • 6. Limitations of the Study
  • 7. Conclusions
  • Acknowledgments
  • Author Note
  • References
  • Part III AI in the Arts and Digital Humanities
  • 13 AI and Digital Innovation in Art History, Archaeology, and Cultural Heritage (Livia P. Bettero)
  • 1. Getting Data
  • 1.1. Open access and open science
  • 1.2. Language corpora
  • 1.3. Web scraping techniques
  • 1.4. Text cleaning and parsing techniques
  • 1.5. Tools for extracting and cleaning data
  • 2. Finding Inspiration in data
  • 2.1. Corpora comparison
  • 2.2. Entity recognition
  • 2.3. Summarization
  • 2.4. Profiling
  • 3. Telling A Story with Data
  • 3.1. AI-generated infographics
  • 3.2. AI-generated animated videos
  • 4. Final Considerations
  • Author Note
  • References
  • 14 Augmented Reality, Past–Present–Future: A Digital Humanities Project (Roberta Rodrigues, Tânia Cerqueira, and Luciano Moreira)
  • 1. Introduction
  • 2. Literature Review
  • 3. Methods
  • 4. Results
  • 5. Discussion
  • Funding
  • Author Note
  • References
  • 15 Using Generative AI and LLMs to Enrich Art Collection Metadata for Searching, Browsing, and Studying Art History in Digital Humanities (Annastiina Ahola, Lilli Peura, Rafael Leal, Heikki Rantala, and Eero Hyvönen)
  • 1. Introduction
  • 2. Related Works
  • 3. Methods for Creating Subject Matter Keywords
  • 4. Comparing GAI tools for Keyword Extraction
  • 4.1. Overview of general performance
  • 4.2. Zooming In: Case-by-Case Examination
  • 4.3. Using textual descriptions
  • 5. Enriching ArtSampo KG and UI using LLMs
  • 6. Discussion and future work
  • Acknowledgments
  • Author Note
  • References
  • 16 The Possibility of the Unpredictability in the Computational Art (Teófilo Augusto da Silva and Suzete Venturelli)
  • 1. Serendipity and Unpredictability
  • 2. Cybernetic Serendipity and the Works of Chance
  • 3. Unpredictability in Generative Art
  • 4. Beabá – Palavras ao acaso by Waldemar Cordeiro and Giorgio Moscati (1968)
  • 5. Amoreiras by Gilbertto Prado and Grupo Poéticas Digitais (2010)
  • 6. Tijolo Esperto by Media Lab / University of Brasília (UnB) (2009)
  • 7. Esculturas Quânticas by Alexandre Rangel (2018)
  • 8. Conclusion
  • Acknowledgments
  • Author Note
  • References

Figures

Figure 1.1: Relationship between speaking, writing, reading, and listening language skills and DET sub-scores

Figure 1.2: Example of an interactive reading sub-task: “Highlight the Answer”

Figure 1.3: Example of an interactive listening sub-task: “Dialogue Completion”

Figure 1.4: Example of an interactive writing task

Figure 1.5: Example of a C-test task

Figure 1.6: Example of a Yes/No vocabulary task

Figure 1.7: Example of a dictation task

Figure 1.8: Example of a picture description task

Figure 2.1: A bar graph depicting the irregularities in ChatGPT’s stop word frequency count, created by subtracting the lowest word frequency from the highest

Figure 2.2: A line graph depicting the top four most irregular stop words, with a primary focus on the word “de”

Figure 2.3: A line graph depicting the top four most irregular stop words, with a primary focus on the word “e”

Figure 2.4: A line graph illustrating the relatively regular and balanced frequency counts of the stop words “em” and “com”

Figure 2.5: Bar graph depicting the frequency of each repetition count, including cases where word frequency is zero (x = 0)

Figure 2.6: Line graph depicting the variability in word frequency for the absence case (x = 0)

Figure 3.1: Machine translation evaluation scores

Figure 4.1: Ludwig platform conceptual architecture diagram

Figure 4.2: Home page and Text Simplifier prototypes, in Portuguese

Figure 6.1: Examples of SentiLex Input (lemmas)

Figure 6.2: Output of the Sentiment Analysis Function

Figure 6.3: Output of the Word Class Function (noun)

Figure 6.4: Output of the Text Readability Function

Figure 6.5: Output of the Lexical Diversity Function

Figure 6.6: Output of the Lexical Density Function

Figure 6.7: Output of the Lexical Sophistication Function

Figure 8.1: Template of gender prediction. The original sentence contains male-associated terms

Figure 8.2: Research methodology flowchart

Figure 8.3: Sample of original sentence and models predicted sentences

Figure 12.1: Results

Figure 14.1: Research phases: Participatory and non-participatory cycle

Figure 14.2: Augmented reality projection of the town’s square (Artivive App)

Figure 15.1: Example of an art object in the KG

Figure 15.2: Comparison of top 15 keywords: Original human-annotated dataset vs. AI-generated keywords

Figure 15.3: AI-generated keywords for two artworks using different prompts: After Sunset (1882) by Victor Westerholm and Portrait of Jalo Sihtola (1910) by Yrjö Ollila

Figure 15.4: Different keywords for art objects are listed in columns in the table result view

Figure 15.5: The most common keywords generated for art objects can be visualized as, e.g., pie charts, by source

Figure 15.6: Search showing art objects with the keyword snow annotated by either humans or GAI

Figure 16.1: ⼖?≷??, Parametric Genetic Algorithm Evolution, 2019

Figure 16.2: REAS, Casey, Process 18, 2010

Figure 16.3: Waldemar Cordeiro and Giorgio Moscati: Beabá, 1968

Figure 16.4: Gilbertto Prado and Grupo Poéticas Contemporâneas: Amoreiras [Mulberry Trees], 2010

Figure 16.5: Suzete Venturelli, Breno Rocha, Tiago Coelho, and Bruno Ribeiro: Tijolo [Smart Brick]

Figure 16.6: Alexandre Rangel: Esculturas Quânticas [Quantum Sculptures], 2018. Microsoft Kinect Sensor and Software

Tables

Table 1.1: Task types and administration order

Table 3.1: Models evaluated

Table 7.1: Parliament dataset overview

Table 7.2: Overall topic assignment distribution by language (count, column %)

Table 7.3: Overall misinformation terms distribution (column %)

Table 7.4: Number of utterances with misinformation term count (column %)

Table 7.5: GB intentional AI coding on misinformation (count of sub-codes, count of code and sub-code instances by topic assignment)

Table 7.6: AT intentional AI coding on misinformation (count of sub-codes, count of code and sub-code instances by topic assignment)

Table 7.7: PT intentional AI coding on misinformation (count of sub-codes, count of code and sub-code instances by topic assignment)

Table 8.1: Total no of sentences for predictions of missing words in sentence

Table 8.2: Detailed predictions on sentimental sentences

Table 9.1: Models used in the case study

Table 9.2: Number of prompts identified with possible toxicity gender bias using the R4 Target model

Table 9.3: Percentage difference between biased prompt means by assessment attribute

Table 11.1: Comparison of cost structures in Germany and China

Table 11.2: Comparison of practical applications in Germany and China

Table 12.1: Quantification of selected studies

Table 12.2: Initial research results

Details

Pages
XVIII, 234
Publication Year
2026
ISBN (PDF)
9781805843252
ISBN (ePUB)
9781805843269
ISBN (Softcover)
9781803749549
DOI
10.3726/b23492
Language
English
Publication date
2026 (October)
Keywords
Artificial Intelligence Digital Humanities Natural Language Processing Ethics Cultural Heritage Creativity Multiliteracies Machine Learning Human-AI Collaboration
Published
Oxford, Berlin, Bruxelles, Chennai, Lausanne, New York, 2026. xviii, 234 pp., 5 fig. col., 37 fig. b/w, 20 tables.
Product Safety
Peter Lang Group AG

Biographical notes

Silvia Lima Gonçalves Araújo (Volume editor) Micaela Aguiar (Volume editor) Dalila Durães (Volume editor)

Silvia Lima Gonçalves Araújo is an Associate Professor at the University of Minho, specializing in digital humanities. Micaela Aguiar is an Assistant Professor at the Faculty of Arts of the University of Porto. Dalila Durães is an Assistant Professor of Computer Science at the Department of Informatics, University of Minho.

Previous

Title: Humanizing Technology, Vol. III