Handwritten Paleographic Greek Text Recognition: A Century-Based Approach

Proceedings of the 13th Conference on Language Resources and Evaluation (LREC 2022), pages 673–679 Marseille, 20-25 June 2022 © European Language Resources Association (ELRA), licensed under CC-BY-NC-4.0 673 MuLVE, A Multi-Language Vocabulary Evaluation Data Set Anik Jacobsen1, Salar Mohtaj1,2, Sebastian M¨oller1,2 1 Technische Universit¨at Berlin, Berlin, Germany 2 German Research Centre for Artificial Intelligence (DFKI), Labor Berlin, Germany a.jacobsen@campus.tu-berlin.de, {salar.mohtaj, sebastian.moeller}@tu-berlin.de Abstract Vocabulary learning is vital to foreign language learning. Correct and adequate feedback is essential to successful and satisfying vocabulary training. However, many vocabulary and language evaluation systems perform on simple rules and do not account for real-life user learning data. This work introduces Multi-Language Vocabulary Evaluation Data Set (MuLVE), a data set consisting of vocabulary cards and real-life user answers, labeled indicating whether the user answer is correct or incorrect. The data source is user learning data from the Phase6 vocabulary trainer. The data set contains vocabulary questions in German and English, Spanish, and French as target language and is available in four different variations regarding pre-processing and deduplication. We experiment to fine-tune pre-trained BERT language models on the downstream task of vocabulary evaluation with the proposed MuLVE data set. The results provide outstanding results of > 95.5 accuracy and F2-score. The data set is available on the European Language Grid. Keywords: Data Sets, Vocabulary Evaluation, Paraphrase Detection 1. Introduction Vocabulary learning is an essential part of foreign lan- guage learning. Building an extensive vocabulary is necessary to master a foreign language and communi- cate successfully (Alqahtani, 2015). To achieve long- term memory of words and their meaning, repetition and appropriate feedback are crucial (Metcalfe and Ko- rnell, 2007). We introduce a Multi-Language Vocabu- lary Evaluation Data Set (MuLVE), including real-life user vocabulary learning data, aiming to improve vo- cabulary evaluation. phase-6 GmbH1 (hereinafter referred to as Phase6) of- fers a digital education tool in the domain of language learning. Their service is a vocabulary trainer, avail- able for various digital devices, optimizing vocabulary training for long-term memory. Students can study vo- cabulary independently and aligned with the content of their school lessons. Phase6 operates in the German market and focuses on pupils. The area of vocabulary evaluation and training has not been addressed in Natural Language Processing (NLP). Most existing language learning systems operate on simple rules that compare the user’s answer to an exist- ing answer, neglecting potential correct answers such as synonyms or various ways of formatting. These in- flexible systems lead to user frustration and a limited learning experience, resulting in users losing interest in language learning completely. We aim to establish a robust and significant multilin- gual data set for vocabulary evaluation from available user learning data provided by Phase6 to allow for the development and training of more flexible systems for the task of vocabulary evaluation. Our contributions are as follows: 1https://www.phase-6.de/ • Multi-Language Vocabulary Evaluation Data Set (MuLVE): a data set containing different varia- tions of vocabulary cards and real-life user an- swers with a binary label indicating whether the answer …

Content truncated to 500 words