In this study, we proposed a novel Lexicon-based pseudo-labeling method utilizing explainable AI(XAI) approach. Existing approach have a fundamental limitation in their robustness because poor classifier leads to inaccurate soft-labeling, and it lead
to poor classifier repetitively. Meanwhile, we generate the lexicon consists of sentiment word based on the explainability score. Then we calculate the confidence of unlabeled data with lexicon and add them into labeled dataset for the robust pseudo-labeling approach. Our proposed method has three contributions. First, the proposed methodology automatically generates a lexicon based on XAI and performs independent pseudo-labeling, thereby guaranteeing higher performance and robustness compared to the existing one. Second, since lexicon-based pseudo-labeling is performed without re-learning in most of models, time efficiency is considerably increased, and third, the generated high-quality lexicon can be available for sentiment analysis of data from similar domains. The effectiveness and efficiency of our proposed method were verified through quantitative comparison with the existing pseudo-labeling method and qualitative review of the generated lexicon.
Post processing is the most conventional approach for correcting errors that are caused by Optical Character Recognition(OCR) systems. Two steps are usually taken to correct OCR errors: detection and corrections. For the first task, supervised machin
e learning methods have shown state-of-the-art performances. Previously proposed approaches have focused most prominently on combining lexical, contextual and statistical features for detecting errors. In this study, we report a novel system to error detection which is based merely on the n-gram counts of a candidate token. In addition to being simple and computationally less expensive, our proposed system beats previous systems reported in the ICDAR2019 competition on OCR-error detection with notable margins. We achieved state-of-the-art F1-scores for eight out of the ten involved European languages. The maximum improvement is for Spanish which improved from 0.69 to 0.90, and the minimum for Polish from 0.82 to 0.84.
Large-scale multi-modal classification aim to distinguish between different multi-modal data, and it has drawn dramatically attentions since last decade. In this paper, we propose a multi-task learning-based framework for the multimodal classificatio
n task, which consists of two branches: multi-modal autoencoder branch and attention-based multi-modal modeling branch. Multi-modal autoencoder can receive multi-modal features and obtain the interactive information which called multi-modal encoder feature, and use this feature to reconstitute all the input data. Besides, multi-modal encoder feature can be used to enrich the raw dataset, and improve the performance of downstream tasks (such as classification task). As for attention-based multimodal modeling branch, we first employ attention mechanism to make the model focused on important features, then we use the multi-modal encoder feature to enrich the input information, achieve a better performance. We conduct extensive experiments on different dataset, the results demonstrate the effectiveness of proposed framework.
Deep Learning-based NLP systems can be sensitive to unseen tokens and hard to learn with high-dimensional inputs, which critically hinder learning generalization. We introduce an approach by grouping input words based on their semantic diversity to s
implify input language representation with low ambiguity. Since the semantically diverse words reside in different contexts, we are able to substitute words with their groups and still distinguish word meanings relying on their contexts. We design several algorithms that compute diverse groupings based on random sampling, geometric distances, and entropy maximization, and we prove formal guarantees for the entropy-based algorithms. Experimental results show that our methods generalize NLP models and demonstrate enhanced accuracy on POS tagging and LM tasks and significant improvements on medium-scale machine translation tasks, up to +6.5 BLEU points. Our source code is available at https://github.com/abdulrafae/dg.
Within the last few years, the number of Arabic internet users and Arabic online content is in exponential growth. Dealing with Arabic datasets and the usage of non-explicit sentences to express an opinion are considered to be the major challenges in
the field of natural language processing. Hence, sarcasm and sentiment analysis has gained a major interest from the research community, especially in this language. Automatic sarcasm detection and sentiment analysis can be applied using three approaches, namely supervised, unsupervised and hybrid approach. In this paper, a model based on a supervised machine learning algorithm called Support Vector Machine (SVM) has been used for this process. The proposed model has been evaluated using ArSarcasm-v2 dataset. The performance of the proposed model has been compared with other models submitted to sentiment analysis and sarcasm detection shared task.
