Natural Language Processing in Albanian Language
Enda Alidema, Trime Ismajli, Jetmir Gjoni, Eliot Bytyçi · 2025
This paper examines the application and challenges of Natural Language Processing (NLP) in the Albanian language, a less-studied computational - linguistic domain. Despite the rapid advancements in NLP technologies, there is a notable lack of research and tools for Albanian, which limits the inclusion of the Albanian-speaking population in the benefits of Artificial Intelligence (AI) advancements. Through a comprehensive review of recent academic papers, this study analyzes the effectiveness of NLP in areas such as hate speech detection, named entity recognition, and news analysis for the Albanian language. The paper emphasizes the importance of developing NLP resources for Albanian to ensure cultural preservation, social inclusion, and technological advancement. It highlights the potential of models like Bidirectional long short-term memory (BiLSTM) and Bidirectional Encoder Representations from Transformers (BERT) in improving the understanding of Albanian text, which is crucial for applications that have a social impact. The findings emphasize the need for continued research and development to encourage an inclusive digital space that respects linguistic diversity and promotes language equality in AI.