Appìl: An Interface for the Improvement of Lexical Skills

Designed and developed a full-stack adaptive language learning platform integrating computational linguistics, cognitive science, and modern web technologies.

The ongoing evolution of digital education has opened new possibilities for second language learning, yet many existing platforms remain constrained by commercial interests and limited pedagogical depth. Despite the proliferation of language learning applications, few are explicitly grounded in robust second language acquisition (SLA) theory (Ellis, 2008; Krashen, 1982; Long, 1996) or capable of offering authentic, learner-driven adaptation. This research emerges within that gap, responding to the growing need for open, scientifically informed, and pedagogically coherent digital tools that can personalize instruction while preserving methodological rigor. This doctoral study explores the intersection of second language acquisition (SLA), digital pedagogy, and computational linguistics through the design, development, and evaluation of Appìl, an adaptive, web-based language learning platform created by me, under supervision, during my doctoral studies, and situated within the broader field of intelligent computer-assisted language learning (ICALL) (Heift & Schulze, 2007). It integrates theoretical insights from SLA, psychology, and cognitive science (Evans, 2008; Kahneman, 2011; Vygotsky, 1986) with technological innovations drawn from natural language processing (NLP) and linguistic linked open data (LLOD) (Chiarcos et al., 2013). The overarching goal is to demonstrate how open, data-driven systems can enhance learner autonomy and motivation (Deci & Ryan, 2014; Zimmerman, 2002) while maintaining instructional coherence within a structured learning framework. Adopting a mixed-methods approach, the present research project combines theoretical synthesis, technical implementation, and empirical evaluation. It includes a systematic review of SLA theories and digital learning methodologies, the computational design of the Appìl platform, and a field-based evaluation of its pedagogical effectiveness. The system’s architecture merges CEFR-aligned progression (Council of Europe, 2020) with NLP-enabled personalization, providing learners with individualized flashcards, adaptive exercises, and real-time feedback tailored to their evolving proficiency levels. Findings reveal that personalization grounded in SLA theory, rather than purely algorithmic adaptation, significantly enhances learner motivation, vocabulary retention, and engagement. Learners responded positively to flexible and user-driven content pathways, although sustaining participation and addressing digital literacy barriers remained ongoing challenges. The development process, conducted by a single researcher, underscored both the potential and the practical constraints of independent, interdisciplinary platform creation, while also highlighting the centrality of human engagement throughout design and implementation.