Aug 2026· International Conference on Machine Vision and Deep Learning· Vol 14326, pp. 143263D - 143263D-7· 0 citations· 6 references
Engineering
TL;DR
This paper addresses the task of automatic identification of German word forms and constructs an end-to-end model framework based on deep neural networks that employs character-level and subword-level dual-channel feature representations, and combines encoder-decoder architecture, scaled dot-product attention, and positional encoding.
Abstract
In the context of the increasing demand for cross-border communication and multilingual text processing, the complex word forms such as case, number, gender, and verb conjugation in German pose significant challenges for large-scale automatic analysis. Traditional rule-based and dictionary-driven methods struggle to balance accuracy and generalization capabilities. This paper addresses the task of automatic identification of German word forms and constructs an end-to-end model framework based on deep neural networks. It employs character-level and subword-level dual-channel feature representations, combines encoder-decoder architecture, scaled dot-product attention, and positional encoding, and jointly models long-distance dependencies within words and fine-grained morphological patterns. At the same time, it designs multi-task learning objectives incorporating morphological constraints such as case, number, and gender, and completes model training using the Adam optimizer, learning rate decay, and early stopping strategies. Experimental results show that on the standard test set, the word form accuracy rate reaches 0.917, the character accuracy rate reaches 0.966, and the average edit distance drops to 0.28. In robust scenarios such as 10% character perturbation and domain transfer, the word form accuracy rates remain at 0.887 and 0.868 respectively, with the average edit distance remaining at most 0.43, and the performance decline is within an acceptable range. This has engineering application value for the automatic processing of complex morphological languages and the construction of multilingual natural language processing systems.
Current advances in neural network models have improved state-of-the-art performance in natural language processing tasks such as named-entity recognition, sentiment analysis, and machine translation. In particular, neural language models are applied to encode information in word embeddings. These approaches are generally trained on large corpora using semi-supervised learning. Word embeddings encode the syntactic and semantic properties of words as dense vectors. In agglutinative languages such as Turkish, Finnish, and Hungarian, word-embedding construction is challenging because extensive suffixation and polysemy can cause information loss. To overcome these limitations, character n-grams are often preferred for embedding representations. Nevertheless, character n-grams do not guarantee the capture of information in long word sequences. In this study, a method that partitions word sequences according to frequent patterns within a given context is proposed for training a neural language model. In this respect, likelihood- and ranking-based inference are combined with n-gram and syllable partitioning for word-embedding generation from a text corpus. The proposed approach provides a language-agnostic, context-sensitive segmentation mechanism that can complement language processing methods such as lemmatization, morphological analysis, and stemming. For embedding generation, the SkipGram and FastText models are used, and the effects of word partitioning are evaluated using analogy, named-entity recognition, POS tagging, sentiment analysis, and morphological disambiguation datasets for Turkish. The results indicate task-dependent and generally limited improvements over traditional token-based word-embedding extraction. In particular, skip n-gram partitioning produces a substantial improvement over partitioning based on frequent-ngrams, sentencepiece-bpe, sentence-unigram and morfessor. No consistent relationship was observed across tasks between performance and either graph density or the average number of distinct n-grams per sentence.
This study integrates parallel multi-kernel word-level convolutional features into conventional and hybrid deep learning models for Arabic text analysis tasks, providing a systematic within-study assessment of model sensitivity to architecture, preprocessing, and learning-rate selection.
Ahmed I.Taloba, George Samy Rady, Khaled F. Hussain· International Journal of Adv...· 0 citations
This paper presents KazMorphLM, a morpheme-aware language model for Kazakh automatic speech recognition (ASR). Kazakh, a highly agglutinative Turkic language, poses a fundamental challenge for conventional word-level language models, since a single root can generate hundreds of inflected forms through productive suffixation, causing extreme data sparsity. Our objective is to overcome this sparsity by modelling language at the morpheme level. The method combines three components: (1) a rule-based morpheme segmenter built on a fully categorized inventory of 118 suffix entries (175 surface forms) across 12 morphological categories, with vowel-harmony validation and consonant-assimilation rules; (2) a two-level interpolated n-gram architecture coupling a 7-gram morpheme model with a 5-gram word model under Witten-Bell smoothing; and (3) a four-channel rescoring mechanism integrating acoustic, word-level, morpheme-level and vowel-harmony scores. Integrated into a hybrid FastConformer–MMS-1B pipeline, KazMorphLM attains 6.86% word error rate (WER) on the FLEURS test set under N-best rescoring, a 15.2% relative reduction over word-level KenLM rescoring (8.09%). On a live evaluation with 44 native speakers (421 recordings), KazMorphLM significantly outperforms word-level KenLM rescoring (6.01% vs. 7.45% WER; Wilcoxon p<0.001). To our knowledge this is the first morpheme-aware, vowel-harmony-informed language model for Kazakh ASR rescoring, with a methodology transferable to other Turkic languages.
This paper proposes a general optimization framework that combines a vocabulary pruning method with a targeted fine-tuning protocol for MNMT models, and reduces the vocabulary size from over 128,000 to approximately 10,000 tokens, enabling a 60% memory saving without any loss in performance.
Ahmed Amine Aliane, N. Semmar, H. Aliane· 0 citations
This study clarifies the evolution of Chinese word segmentation technology, providing a reference for the selection, engineering implementation, and optimization of word segmentation algorithms in the large model era, and is highly valuable for advancing the high-quality development of Chinese natural language processing.
Yuemeng Ren· International journal of eng...· 0 citations
Alation studies reveal a strong synergy between POS features and CRF decoding, with the combined approach yielding the best single-model performance, and ensembling models trained with different objectives improves both overall F1 score and discontinuous MWE scores, demonstrating the importance of training diversity for capturing non-adjacent syntactic patterns.
Yunus Karatepe, Mert Sülük, Zeynep Tu˘gçe Kırımlı et al.· 0 citations
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
