With recent advances in natural language generation, risks associated with the rapid proliferation and misuse of generative language models for malicious purposes steadily increase. Artificial text detection (ATD) has emerged to develop resources and computational methods to mitigate these risks, such as generating fake news and scientific article reviews. This paper introduces corpus of artificial texts (CoAT), a large-scale corpus of human-written and generated texts for the Russian language. CoAT spans six domains and comprises outputs from 13 text generation models (TGMs), which differ in the number of parameters, architectural choices, pre-training objectives, and downstream applications. We detail the data collection methodology, conduct a linguistic analysis of the corpus, and present a detailed analysis of the ATD experiments with widely used artificial text detectors. The results demonstrate that the detectors perform well on the seen TGMs, but fail to generalise to unseen TGMs and domains. We also find it challenging to identify the author of the given text, and human annotators significantly underperform the detectors. We release CoAT, the codebase, two ATD leaderboards, and other materials used in the paper.

This paper provides data resources for low-resource hate speech detection. Specifically, we introduce two different data resources: (i) the HateBR 2.0 corpus, which is composed of 7,000 comments extracted from Brazilian politicians’ accounts on Instagram and manually annotated a binary class (offensive versus non-offensive) and hate speech targets. It consists of an updated version of the HateBR corpus, in which highly similar and one-word comments were replaced; and (ii) the multilingual offensive lexicon (MOL), which consists of 1,000 explicit and implicit terms and expressions annotated with context information. The lexicon also comprises native-speaker translations and its cultural adaptations in English, Spanish, French, German, and Turkish. Both corpus and lexicon were annotated by three different experts and achieved high inter-annotator agreement. Lastly, we implemented baseline experiments on the proposed data resources. Results demonstrate the reliability of data outperforming baseline dataset results in Portuguese, besides presenting promising results for hate speech detection in different languages.

Intent detection and slot filling are two necessary tasks for natural language understanding. Deep neural models have already shown great ability facing sequence labeling and sentence classification tasks, but they require a large amount of training data to achieve accurate results. However, in many low-resource languages, creating accurate training data is problematic. Consequently, in most of the language processing tasks, low-resource languages have significantly lower accuracy than rich-resource languages. Hence, training models in low-resource languages with data from a richer-resource language can be advantageous. To solve this problem, in this paper, we used pretrained language models, namely multilingual BERT (mBERT) and XLM-RoBERTa, in different cross-lingual and monolingual scenarios. To evaluate our proposed model, we translated a small part of the Airline Travel Information System (ATIS) dataset into Persian. Furthermore, we repeated the experiments on the MASSIVE dataset to increase our results’ reliability. Experimental results on both datasets show that the cross-lingual scenarios significantly outperform monolinguals ones.

Part of speech (POS) tagging, though considered to be preliminary to any Natural Language Processing (NLP) task, is crucial to account for, especially in low resource language like Khasi that lacks any form of formal corpus. POS tagging is context sensitive. Therefore, the task is challenging. In this paper, we attempt to investigate a deep learning approach to the POS tagging problem in Khasi. A deep learning model called Robustly Optimized BERT Pretraining Approach (RoBERTa) is pretrained for language modelling task. We then create RoBERTa for POS (RoPOS) tagging, a model that performs POS tagging by fine-tuning the pretrained RoBERTa and leveraging its embeddings for downstream POS tagging. The existing tagset that has been designed, customarily, for the Khasi language is employed for this work, and the corresponding tagged dataset is taken as our base corpus. Further, we propose additional tags to this existing tagset to meet the requirements of the language and have increased the size of the existing Khasi POS corpus. Other machine learning and deep learning models have also been tried and tested for the same task, and a comparative analysis is made on the various models employed. Two different setups have been used for the RoPOS model, and the best testing accuracy achieved is 92 per cent. Comparative analysis of RoPOS with the other models indicates that RoPOS outperforms the others when used for inferencing on texts that are outside the domain of the POS tagged training dataset.

There has been a steep rise in user-generated content on the Web and social media platforms during the last few years. While the ease of content creation allows anyone to create content, at the same time it is difficult to monitor and control the spread of detrimental content. Recent research in natural language processing and machine learning has shown some hope for the purpose. Approaches and methods are now being developed for the automatic flagging of problematic textual content, namely hate speech, cyberbullying, or fake news, though mostly for English language texts. This paper presents an algorithmic approach based on deep learning models for the detection of violent incidents from tweets in the Spanish language (binary classification) and categorizes them further into five classes – accident, homicide, theft, kidnapping, and none (multi-label classification). The performance is evaluated on the recently shared benchmark dataset, and it is found that the proposed approach outperforms the various deep learning models, with a weighted average precision, recall, and F1-score of 0.82, 0.81, and 0.80, respectively, for the binary classification. Similarly, for the multi-label classification, the proposed model reports weighted average precision, recall, and F1-score of 0.54, 0.79, and 0.64, respectively, which is also superior to the existing results reported in the literature. The study, thus, presents meaningful contribution to detection of violent incidents in Spanish language social media posts.

Datasets serve as crucial training resources and model performance trackers. However, existing datasets have exposed a plethora of problems, inducing biased models and unreliable evaluation results. In this paper, we propose a model-agnostic dataset evaluation framework for automatic dataset quality evaluation. We seek the statistical properties of the datasets and address three fundamental dimensions: reliability, difficulty, and validity, following a Classical Test Theory (CTT). Taking the named entity recognition (NER) datasets as a case study, we introduce nine statistical metrics for a statistical dataset evaluation framework. Specifically, we investigate the reliability of a NER dataset with three metrics, including Redundancy, Accuracy, and Leakage Ratio. We assess the dataset difficulty through four metrics: Unseen Entity Ratio, Entity Ambiguity Degree, Entity Density, and Model Differentiation. For validity, we introduce the Entity Imbalance Degree and Entity-Null Rate to evaluate the effectiveness of the dataset in assessing language model performance. Experimental results validate that our evaluation framework effectively assesses various aspects of the dataset quality. Furthermore, we study how the dataset scores on our statistical metrics affect the model performance and appeal for dataset quality evaluation or targeted dataset improvement before training or testing models.

When humans solve complex problems, they typically construct, reflect, and revise sequences of ideas, hypotheses, and beliefs until a final decision or conclusion is reached. Contrary to this, current machine learning models are mostly trained to map an input to one single and fixed output. In this paper, we investigate how we can equip models with the ability to represent, construct, and evaluate a second, third, and $k$-th thought within their prediction process. Drawing inspiration from Hegel’s dialectics, we propose and evaluate the thought flow concept which constructs a sequence of predictions. We present a self-correction mechanism which (a) is trained to estimate the model’s correctness and which (b) performs iterative prediction updates based on the gradient of the correctness prediction. We introduce our method focusing initially on question answering (QA) and carry out extensive experiments which demonstrate that (i) our method is able to correct its own predictions and that (ii) it can improve model performance by a large margin. In addition, we conduct a qualitative analysis of thought flow correction patterns and explore how thought flow predictions affect users’ human-AI collaboration in a crowdsourcing study. We find that (iii) thought flows improve user performance and are perceived as more natural, correct, and intelligent regarding single and/or top-3 predictions.

In spite of the superior performance deep neural networks have proven in thousands of applications in the past few years, addressing the over-sensitivity of these models to noise and/or intentional slight perturbations is still an active area of research. In the computer vision domain, perturbations can be directly applied to the input images. The task in the natural language processing domain is quite harder due to the discrete nature of natural languages. There has been a considerable amount of effort put to address this problem in high-resource languages like English. However, there is still an apparent lack of such studies in the Arabic language, and we aim to be the first to conduct such a study in this work. In this study, we start by training seven different models on a sentiment analysis task. Then, we propose a method to attack our models by means of the worst synonym replacement where the synonyms are automatically selected via the gradients of the input representations. After proving the effectiveness of the proposed adversarial attack, we aim to design a framework that enables the development of models robust to attacks. Three different frameworks are proposed in this work and a thorough comparison between the performance of these frameworks is presented. The three scenarios revolve around training the proposed models either on adversarial samples only or also including clean samples beside the adversarial ones, and whether or not to include weight perturbation during training.

Semantic representation is the task of conveying the meaning of a natural language utterance by converting it to a logical form that can be processed and understood by machines. It is utilised by many applications in natural language processing (NLP), particularly in tasks relevant to natural language understanding (NLU). Due to the widespread use of semantic parsing in NLP, many semantic representation schemes with different forms have been proposed; Universal Conceptual Cognitive Annotation (UCCA) is one of them. UCCA is a cross-lingual semantic annotation framework that allows easy annotation without requiring substantial linguistic knowledge. UCCA-annotated datasets have been released so far for English, French, German, Russian, and Hebrew. In this paper, we present a UCCA-annotated Turkish dataset of 400 sentences that are obtained from the METU-Sabanci Turkish Treebank. We provide the UCCA annotation specifications defined for the Turkish language so that it can be extended further. We followed a semi-automatic annotation approach, where an external semantic parser is utilised for the initial annotation of the dataset, which is manually revised by two annotators. We used the same semantic parser model to evaluate the dataset with zero-shot and few-shot learning, demonstrating that even a small sample set from the target language in the training data has a notable impact on the performance of the parser (15.6% and 2.5% gain over zero-shot for labelled and unlabelled results, respectively).

Warning: This paper is based on hate speech detection and may contain examples of abusive/ offensive phrases.

Cyberbullying, online harassment, etc., via offensive comments are pervasive across different social media platforms like ™Twitter, ™Facebook, ™YouTube, etc. Hateful comments must be detected and eradicated to prevent harassment and violence on social media. In the Natural Language Processing (NLP) domain, the most prevalent task is comment classification, which is challenging, and language models based on transformers are at the forefront of this advancement. This paper intends to analyze the performance of language models based on transformers like BERT, ALBERT, RoBERTa, and DistilBERT on the Indian hate speech datasets over binary classification. Here, we utilize the existing datasets, i.e., HASOC (Hindi and Marathi) and HS-Bangla. So, we evaluate several multilingual language models like MuRIL-BERT, XLM-RoBERTa, etc., few monolingual language models like RoBERTa-Hindi, Maha-BERT (Marathi), Bangla-BERT (Bangla), Assamese-BERT (Assamese), etc., and perform cross-lingual experiment also. For further analyses, we perform multilingual, monolingual, and cross-lingual experiments on our Hate Speech Assamese (HS-Assamese) (Indo-Aryan language family) and Hate Speech Bodo (HS-Bodo) (Sino-Tibetan language family) dataset (HS dataset version 2) also and achieved a promising result. The motivation of the cross-lingual experiment is to encourage researchers to learn about the power of the transformer. Note that no pre-trained language models are currently available for Bodo or any other Sino-Tibetan languages.

This paper presents XeraWord, an innovative tool for automatically generating nominal phrases. XeraWord can be used for different tasks, ranging from teaching languages to the creation of examples in lexicography, or even for the development of resources for natural language processing. In this area, Xera was the first experiment, allowing the automatic generation of nominal phrases in three languages: German, French and Spanish. This tool was extended to support other languages, namely, Portuguese and Galician.

We start by presenting the theory behind the development of Xera and its new version, XeraWord, namely, the applied base methodology, and the natural language processing resources used to support it. Then, TraduWord, a tool specifically developed to construct resources for new languages, is presented. This tool allows the semi-automatic translation of the data required for the nominal phrase generation. For this, we discuss its advantages and disadvantages, analysing the quality of the translated resources, as well as the amount of manual work required to validate and correct these resources.

Human evaluation of machine translation is considered the “gold standard” for evaluation, but it remains a challenging task for which to define best practices. Recent work has focused on incorporating intersentential context into human evaluation, to better distinguish between high-performing machine translation systems and human translations. In this work, we examine several ways that such context influences evaluation and evaluation protocols. We take a close look at annotator variation through the lens of calibration sets and focus on the implications for context-aware evaluation protocols. We then demonstrate one way in which degraded target-side intersentential context can influence annotator scores of individual sentences, a finding that supports the context-aware approach to evaluation and which also has implications for best practices in evaluation protocols.

Decompounding is an essential preprocessing step in text-processing tasks such as machine translation, speech recognition, and information retrieval (IR). Here, the IR issues are explored from five viewpoints. (A) Does word decompounding impact the Indian language IR? If yes, to what extent? (B) Can corpus-based decompounding models be used in the Indian language IR? If yes, how? (C) Can machine learning and deep learning-based decompounding models be applied in the Indian language IR? If yes, how? (D) Among the different decompounding models (corpus-based, hybrid machine learning-based, and deep learning-based), which provides the best effectiveness in the IR domain? (E) Among the different IR models, which provides the best effectiveness from the IR perspective? This study proposes different corpus-based, hybrid machine learning-based, and deep learning-based decompounding models in Indian languages (Marathi, Hindi, and Sanskrit). Moreover, we evaluate the effectiveness of each activity from an IR perspective only. It is observed that the different decompounding models improve IR effectiveness. The deep learning-based decompounding models outperform the corpus-based and hybrid machine learning-based models in Indian language IR. Among the different deep learning-based models, the Bi-LSTM-A model performs best and improves mean average precision (MAP) by 28.02% in Marathi. Similarly, the Bi-RNN-A model improves MAP by 18.18% and 6.1% in Hindi and Sanskrit, respectively. Among the retrieval models, the In_expC2 model outperforms others in Marathi and Hindi, and the BB2 model outperforms others in Sanskrit.

Multilingual pre-trained models make it possible to develop natural language processing (NLP) applications for low-resource languages (LRLs) using the model of resource-rich languages (RRLs). However, the structural characteristics of the target languages can impact task-specific learning. In this paper, we investigate the influence of structural diversities of languages on the system’s overall performance. Specifically, we propose a customized approach to leverage task-specific data of low-resource language families via transfer learning from RRL. Our findings are based on question-answering tasks using the XLM-R, mBERT, and IndicBERT transformer models and Indic languages (Hindi, Bengali, and Telugu). On the XQuAD-Hindi dataset, the few-shot learning using Bengali improves the benchmark mBERT (F1/EM) score by +(10.86/7.87) and XLM-R score by +(3.84/4.42). Few-shot learning using Telugu has also improved the mBERT score by +(10.42/7.36) and +(3.04/2.72) for XLM-R. In addition, our model has demonstrated benchmark-compatible performance in a zero-shot setup with single-epoch task learning. This approach can be adapted for other NLP tasks for LRLs.

Language processing systems such as part-of-speech (POS) tagging, named entity recognition, machine translation, speech recognition, and language modeling have been well-studied in high-resource languages. Nevertheless, research on these systems for several low-resource languages, including Bodo, Mizo, Nagamese, and others, is either yet to commence or is in its nascent stages. Language model (LM) plays a vital role in the downstream tasks of modern natural language processing. Extensive studies are carried out on LMs for high-resource languages. However, these low-resource languages are still underreprese. In this study, we first present BodoBERT, an LM for the Bodo language. To the best of our knowledge, this work is the first such effort to develop an LM for Bodo. Second, we present an ensemble deep learning-based POS tagging model for Bodo. The POS tagging model is based on combinations of BiLSTM with conditional random field and stacked embedding of BodoBERT with BytePairEmbeddings. We cover several LMs in the experiment to see how well they work in POS tagging tasks. The best-performing model achieves an F1 score of 0.8041. A comparative experiment was also conducted on Assamese POS taggers, considering that the language is spoken in the same region as Bodo.

Statistical Machine Translation (SMT) systems use various probabilistic and statistical Natural Language Processing (NLP) methods to automatically translate from one language to another language while retaining the originality of the context. This paper aims to discuss the development of bilingual SMT models for translating English into fifteen low-resource Indic languages (ILs) and vice versa. The process to build the SMT model is described and explained using a workflow diagram. Samanantar and OPUS corpus are utilized for training, and Flores200 corpus is used for fine-tuning and testing purposes. The paper also highlights various preprocessing methods used to deal with corpus noise. The Moses open-source SMT toolkit is being investigated for the system’s development. The impact of distance-based reordering and Morpho-syntactic Descriptor Bidirectional Finite-State Encoder (msd-bidirectional-fe) reordering on ILs is compared in the paper. This paper provides a comparison of SMT models with Neural Machine Translation (NMT) for ILs. All the experiments assess the translation quality using standard metrics such as BiLingual Evaluation Understudy, Rank-based Intuitive Bilingual Evaluation Score, Translation Edit Rate, and Metric for Evaluation of Translation with Explicit Ordering. From the result, it is observed that msd-bidirectional-fe reordering performs better than the distance-based reordering model for ILs. It is also noticed that even though the IL-English and English-IL systems are trained using the same corpus, the former performs better for all the evaluation metrics. The comparison between SMT and NMT shows that across various languages, SMT performs better in some cases, while NMT outperforms in others.

Morphological analysis is a crucial preprocessing stage for building the state-of-the-art natural language processing applications. We propose a bidirectional LSTM (long short-term memory)-based approach to develop the morphological analyzer for the Gujarati language. Our morph analyzer predicts a root word and the morphological features for the given inflected word. We have experimented with two different methods for label representation for predicting morphological features: the monolithic representation method and the individual label representation method. We have also created the gold morphological dataset of 16,234 unique words for the Gujarati language. The dataset contains morpheme splitting and grammatical feature information for each inflected word. Due to the change in the label representation technique in the proposed model, the accuracy of the present baseline system is improved by a large margin. The proposed system performs very well across the POS categories without the knowledge of language-specific suffix rules.

Language identification (LID) is a crucial preliminary process in the field of Automatic Speech Recognition (ASR) that involves the identification of a spoken language from audio samples. Contemporary systems that can process speech in multiple languages require users to expressly designate one or more languages prior to utilisation. The LID task assumes a significant role in scenarios where ASR systems are unable to comprehend the spoken language in multilingual settings, leading to unsuccessful speech recognition outcomes. The present study introduces convolutional recurrent neural network (CRNN)-based LID, designed to operate on the mel-frequency cepstral coefficient (MFCC) characteristics of audio samples. Furthermore, we replicate certain state-of-the-art methodologies, specifically the convolutional neural network (CNN) and Attention-based convolutional recurrent neural network (CRNN with Attention), and conduct a comparative analysis with our CRNN-based approach. We conducted comprehensive evaluations on thirteen distinct Indian languages, and our model resulted in over 98 per cent classification accuracy. The LID model exhibits high-performance levels ranging from 97 per cent to 100 per cent for languages that are linguistically similar. The proposed LID model exhibits a high degree of extensibility to additional languages and demonstrates a strong resistance to noise, achieving 91.2 per cent accuracy in a noisy setting when applied to a European Language (EU) dataset.

Sarcasm detection research in the Bengali language so far can be considered to be narrow due to the unavailability of resources. In this paper, we introduce a large-scale self-annotated Bengali corpus for sarcasm detection research problem in the Bengali language named ‘Ben-Sarc’ containing 25,636 comments, manually collected from different public Facebook pages and evaluated by external evaluators. Then we present a complete strategy to utilize different models of traditional machine learning, deep learning, and transfer learning to detect sarcasm from text using the Ben-Sarc corpus. Finally, we demonstrate a comparison between the performance of traditional machine learning, deep learning, and transfer learning models on our Ben-Sarc corpus. Transfer learning using Indic-Transformers Bengali Bidirectional Encoder Representations from Transformers as a pre-trained source model has achieved the highest accuracy of 75.05%. The second-highest accuracy is obtained by the long short-term memory model with 72.48% and Multinomial Naive Bayes is acquired the third highest with 72.36% accuracy for deep learning and machine learning, respectively. The Ben-Sarc corpus is made publicly available in the hope of advancing the Bengali Natural Language Processing Community. The Ben-Sarc is available at https://github.com/sanzanalora/Ben-Sarc.