Existing text representations such as embeddings and bag-of-words are not suitable for rule learning due to their high dimensionality and absent or questionable feature-level interpretability. This article explores whether large language models (LLMs) could address this by extracting a small number of interpretable features from text. We demonstrate this process on two datasets (CORD-19 and M17+) containing several thousand scientific articles from multiple disciplines and a target being a proxy for research impact. An evaluation based on testing for the statistically significant correlation with research impact has shown that LLama 2-generated features are semantically meaningful. We consequently used these generated features in text classification to predict the binary target variable representing the citation rate for the CORD-19 dataset and the ordinal 5-class target representing an expert-awarded grade in the M17+ dataset. Machine-learning models trained on the LLM-generated features provided similar predictive performance to the state-of-the-art embedding model SciBERT for scientific text. The LLM used only 62 features compared to 768 features in SciBERT embeddings, and these features were directly interpretable, corresponding to notions such as article methodological rigor, novelty, or grammatical correctness. As the final step, we extract a small number of well-interpretable action rules. Consistently competitive results obtained with the same LLM feature set across both thematically diverse datasets show that this approach generalizes across domains.
NLP datasets may still contain annotation errors, even when they are manually annotated. Researchers have attempted to develop methods to automatically reduce the adverse effect of errors in datasets. However, existing methods are time-consuming because they require many trained models to detect errors. This paper proposes a time-saving method that utilizes a tokenization technique called subword regularization to simulate multiple error detection models for detecting errors. Our proposed method, SubRegWeigh, can perform annotation weighting four to five times faster than the existing method. Additionally, SubRegWeigh improved performance in document classification and named entity recognition tasks. In experiments with pseudo-incorrect labels, SubRegWeigh clearly identifies pseudo-incorrect labels as annotation errors. Our code is available at https://github.com/4ldk/SubRegWeigh .
Recent advances in language modelling has significantly decreased the need of labelled data in text classification tasks. Transformer-based models, pre-trained on unlabeled data, can outmatch the performance of models trained from scratch for each task. However, the amount of labelled data need to fine-tune such type of model is still considerably high for domains requiring expert-level annotators, like the legal domain. This paper investigates the best strategies for optimizing the use of a small labeled dataset and large amounts of unlabeled data and perform a classification task in the legal area with 50 predefined topics. More specifically, we use the records of demands to a Brazilian Public Prosecutor's Office aiming to assign the descriptions in one of the subjects, which currently demands deep legal knowledge for manual filling. The task of optimizing the performance of classifiers in this scenario is especially challenging, given the low amount of resources available regarding the Portuguese language, especially in the legal domain. Our results demonstrate that classic supervised models such as logistic regression and SVM and the ensembles random forest and gradient boosting achieve better performance along with embeddings extracted with word2vec when compared to BERT language model. The latter demonstrates superior performance in association with the architecture of the model itself as a classifier, having surpassed all previous models in that regard. The best result was obtained with Unsupervised Data Augmentation (UDA), which jointly uses BERT, data augmentation, and strategies of semi-supervised learning, with an accuracy of 80.7% in the aforementioned task.
Information Extraction (IE) and Text Classification (CLS) serve as the fundamental pillars of NLU, with both disciplines relying on analyzing input sequences to categorize outputs into pre-established schemas. However, there is no existing encoder-based model that can unify IE and CLS tasks from this perspective. To fully explore the foundation shared within NLU tasks, we have proposed a Recursive Method with Explicit Schema Instructor for Universal NLU. Specifically, we firstly redefine the true universal information extraction (UIE) with a formal formulation that covers almost all extraction schemas, including quadruples and quintuples which remain unsolved for previous UIE models. Then, we expands the formulation to all CLS and multi-modal NLU tasks. Based on that, we introduce RexUniNLU, an universal NLU solution that employs explicit schema constraints for IE and CLS, which encompasses all IE and CLS tasks and prevent incorrect connections between schema and input sequence. To avoid interference between different schemas, we reset the position ids and attention mask matrices. Extensive experiments are conducted on IE, CLS in both English and Chinese, and multi-modality, revealing the effectiveness and superiority. Our codes are publicly released.
In the field of natural language processing, text classification, as a basic task, has important research value and application prospects. Traditional text classification methods usually rely on feature representations such as the bag of words model or TF-IDF, which overlook the semantic connections between words and make it challenging to grasp the deep structural details of the text. Recently, GNNs have proven to be a valuable asset for text classification tasks, thanks to their capability to handle non-Euclidean data efficiently. However, the existing text classification methods based on GNN still face challenges such as complex graph structure construction and high cost of model training. This paper introduces a text classification optimization algorithm utilizing graph neural networks. By introducing adaptive graph construction strategy and efficient graph convolution operation, the accuracy and efficiency of text classification are effectively improved. The experimental results demonstrate that the proposed method surpasses traditional approaches and existing GNN models across multiple public datasets, highlighting its superior performance and feasibility for text classification tasks.
With the evolution of generative linguistic steganography techniques, conventional steganalysis falls short in robustly quantifying the alterations induced by steganography, thereby complicating detection. Consequently, the research paradigm has pivoted towards deep-learning-based linguistic steganalysis. This study offers a comprehensive review of existing contributions and evaluates prevailing developmental trajectories. Specifically, we first provided a formalized exposition of the general formulas for linguistic steganalysis, while comparing the differences between this field and the domain of text classification. Subsequently, we classified the existing work into two levels based on vector space mapping and feature extraction models, thereby comparing the research motivations, model advantages, and other details. A comparative analysis of the experiments is conducted to assess the performances. Finally, the challenges faced by this field are discussed, and several directions for future development and key issues that urgently need to be addressed are proposed.
Hierarchical Text Classification (HTC) aims to categorize text data based on a structured label hierarchy, resulting in predicted labels forming a sub-hierarchy tree. The semantics of the text should align with the semantics of the labels in this sub-hierarchy. With the sub-hierarchy changing for each sample, the dynamic nature of text-label alignment poses challenges for existing methods, which typically process text and labels independently. To overcome this limitation, we propose a Text-Label Alignment (TLA) loss specifically designed to model the alignment between text and labels. We obtain a set of negative labels for a given text and its positive label set. By leveraging contrastive learning, the TLA loss pulls the text closer to its positive label and pushes it away from its negative label in the embedding space. This process aligns text representations with related labels while distancing them from unrelated ones. Building upon this framework, we introduce the Hierarchical Text-Label Alignment (HTLA) model, which leverages BERT as the text encoder and GPTrans as the graph encoder and integrates text-label embeddings to generate hierarchy-aware representations. Experimental results on benchmark datasets and comparison with existing baselines demonstrate the effectiveness of HTLA for HTC.
Backdoors can be injected into NLP models to induce misbehavior when the input text contains a specific feature, known as a trigger, which the attacker secretly selects. Unlike fixed words, phrases, or sentences used in the static text trigger, NLP dynamic backdoor attacks design triggers associated with abstract and latent text features, making them considerably stealthier than traditional static backdoor attacks. However, existing research on NLP backdoor detection primarily focuses on defending against static backdoor attacks, while detecting dynamic backdoors in NLP models remains largely unexplored. This paper presents CLIBE, the first framework to detect dynamic backdoors in Transformer-based NLP models. CLIBE injects a "few-shot perturbation" into the suspect Transformer model by crafting optimized weight perturbation in the attention layers to make the perturbed model classify a limited number of reference samples as a target label. Subsequently, CLIBE leverages the generalization ability of this few-shot perturbation to determine whether the original model contains a dynamic backdoor. Extensive evaluation on three advanced NLP dynamic backdoor attacks, two widely-used Transformer frameworks, and four real-world classification tasks strongly validates the effectiveness of CLIBE. We also demonstrate the robustness of CLIBE against various adaptive attacks. Furthermore, we employ CLIBE to scrutinize 49 popular Transformer models on Hugging Face and discover one exhibiting a high probability of containing a dynamic backdoor. We have contacted Hugging Face and provided detailed evidence of this model's backdoor behavior. Moreover, we extend CLIBE to detect backdoor text generation models modified to exhibit toxic behavior. To the best of our knowledge, CLIBE is the first framework capable of detecting backdoors in text generation models without access to trigger input test samples.
The relentless expansion of deep learning applications in recent years has prompted a pivotal shift toward on-device execution, driven by the urgent need for real-time processing, heightened privacy concerns, and reduced latency across diverse domains. This article addresses the challenges inherent in optimising the execution of deep neural networks (DNNs) on mobile devices, with a focus on device heterogeneity, multi-DNN execution, and dynamic runtime adaptation. We introduce CARIn, a novel framework designed for the optimised deployment of both single- and multi-DNN applications under user-defined service-level objectives. Leveraging an expressive multi-objective optimisation framework and a runtime-aware sorting and search algorithm (RASS) as the MOO solver, CARIn facilitates efficient adaptation to dynamic conditions while addressing resource contention issues associated with multi-DNN execution. Notably, RASS generates a set of configurations, anticipating subsequent runtime adaptation, ensuring rapid, low-overhead adjustments in response to environmental fluctuations. Extensive evaluation across diverse tasks, including text classification, scene recognition, and face analysis, showcases the versatility of CARIn across various model architectures, such as Convolutional Neural Networks and Transformers, and realistic use cases. We observe a substantial enhancement in the fair treatment of the problem's objectives, reaching 1.92x when compared to single-model designs and up to 10.69x in contrast to the state-of-the-art OODIn framework. Additionally, we achieve a significant gain of up to 4.06x over hardware-unaware designs in multi-DNN applications. Finally, our framework sustains its performance while effectively eliminating the time overhead associated with identifying the optimal design in response to environmental challenges.
Recent advancements in large language models (LLMs) have opened new avenues for enhancing text classification efficiency in political science, surpassing traditional machine learning methods that often require extensive feature engineering, human labeling, and task-specific training. However, their effectiveness in achieving high classification accuracy remains questionable. This paper introduces a three-stage in-context learning approach that leverages LLMs to improve classification accuracy while minimizing experimental costs. Our method incorporates automatic enhanced prompt generation, adaptive exemplar selection, and a consensus mechanism that resolves discrepancies between two weaker LLMs, refined by an advanced LLM. We validate our approach using datasets from the BBC news reports, Kavanaugh Supreme Court confirmation, and 2018 election campaign ads. The results show significant improvements in classification F1 score (+0.36 for zero-shot classification) with manageable economic costs (-78% compared with human labeling), demonstrating that our method effectively addresses the limitations of traditional machine learning while offering a scalable and reliable solution for text analysis in political science.
This paper presents the experiments and results for the CheckThat! Lab at CLEF 2024 Task 6: Robustness of Credibility Assessment with Adversarial Examples (InCrediblAE). The primary objective of this task was to generate adversarial examples in five problem domains in order to evaluate the robustness of widely used text classification methods (fine-tuned BERT, BiLSTM, and RoBERTa) when applied to credibility assessment issues. This study explores the application of ensemble learning to enhance adversarial attacks on natural language processing (NLP) models. We systematically tested and refined several adversarial attack methods, including BERT-Attack, Genetic algorithms, TextFooler, and CLARE, on five datasets across various misinformation tasks. By developing modified versions of BERT-Attack and hybrid methods, we achieved significant improvements in attack effectiveness. Our results demonstrate the potential of modification and combining multiple methods to create more sophisticated and effective adversarial attack strategies, contributing to the development of more robust and secure systems.
This paper examines the performance of two Large Language Models (LLMs), GPT3.5 and Llama2 and one Small Language Model (SLM) Gemma, across three different classification tasks within the climate change (CC) and environmental domain. Employing BERT-based models as a baseline, we compare their efficacy against these transformer-based models. Additionally, we assess the models' self-evaluation capabilities by analyzing the calibration of verbalized confidence scores in these text classification tasks. Our findings reveal that while BERT-based models generally outperform both the LLMs and SLM, the performance of the large generative models is still noteworthy. Furthermore, our calibration analysis reveals that although Gemma is well-calibrated in initial tasks, it thereafter produces inconsistent results; Llama is reasonably calibrated, and GPT consistently exhibits strong calibration. Through this research, we aim to contribute to the ongoing discussion on the utility and effectiveness of generative LMs in addressing some of the planet's most urgent issues, highlighting their strengths and limitations in the context of ecology and CC.
Electronic Health Records are large repositories of valuable clinical data, with a significant portion stored in unstructured text format. This textual data includes clinical events (e.g., disorders, symptoms, findings, medications and procedures) in context that if extracted accurately at scale can unlock valuable downstream applications such as disease prediction. Using an existing Named Entity Recognition and Linking methodology, MedCAT, these identified concepts need to be further classified (contextualised) for their relevance to the patient, and their temporal and negated status for example, to be useful downstream. This study performs a comparative analysis of various natural language models for medical text classification. Extensive experimentation reveals the effectiveness of transformer-based language models, particularly BERT. When combined with class imbalance mitigation techniques, BERT outperforms Bi-LSTM models by up to 28% and the baseline BERT model by up to 16% for recall of the minority classes. The method has been implemented as part of CogStack/MedCAT framework and made available to the community for further research.
Text classification is crucial for applications such as sentiment analysis and toxic text filtering, but it still faces challenges due to the complexity and ambiguity of natural language. Recent advancements in deep learning, particularly transformer architectures and large-scale pretraining, have achieved inspiring success in NLP fields. Building on these advancements, this thesis explores three challenging settings in text classification by leveraging the intrinsic knowledge of pretrained language models (PLMs). Firstly, to address the challenge of selecting misleading yet incorrect distractors for cloze questions, we develop models that utilize features based on contextualized word representations from PLMs, achieving performance that rivals or surpasses human accuracy. Secondly, to enhance model generalization to unseen labels, we create small finetuning datasets with domain-independent task label descriptions, improving model performance and robustness. Lastly, we tackle the sensitivity of large language models to in-context learning prompts by selecting effective demonstrations, focusing on misclassified examples and resolving model ambiguity regarding test example labels.
Large Language Models have undoubtedly revolutionized the Natural Language Processing field, the current trend being to promote one-model-for-all tasks (sentiment analysis, translation, etc.). However, the statistical mechanisms at work in the larger language models struggle to exploit the relevant information when it is very sparse, when it is a weak signal. This is the case, for example, for the classification of long domain-specific documents, when the relevance relies on a single relevant word or on very few relevant words from technical jargon. In the medical domain, it is essential to determine whether a given report contains critical information about a patient's condition. This critical information is often based on one or few specific isolated terms. In this paper, we propose a hierarchical model which exploits a short list of potential target terms to retrieve candidate sentences and represent them into the contextualized embedding of the target term(s) they contain. A pooling of the term(s) embedding(s) entails the document representation to be classified. We evaluate our model on one public medical document benchmark in English and on one private French medical dataset. We show that our narrower hierarchical model is better than larger language models for retrieving relevant long documents in a domain-specific context.
Generative large language models (LLMs) have achieved state-of-the-art results on a wide range of tasks, yet they remain susceptible to backdoor attacks: carefully crafted triggers in the input can manipulate the model to produce adversary-specified outputs. While prior research has predominantly focused on backdoor risks in vision and classification settings, the vulnerability of LLMs in open-ended text generation remains underexplored. To fill this gap, we introduce BackdoorLLM (Our BackdoorLLM benchmark was awarded First Prize in the SafetyBench competition, https://www.mlsafety.org/safebench/winners, organized by the Center for AI Safety, https://safe.ai/.), the first comprehensive benchmark for systematically evaluating backdoor threats in text-generation LLMs. BackdoorLLM provides: (i) a unified repository of benchmarks with a standardized training and evaluation pipeline; (ii) a diverse suite of attack modalities, including data poisoning, weight poisoning, hidden-state manipulation, and chain-of-thought hijacking; (iii) over 200 experiments spanning 8 distinct attack strategies, 7 real-world scenarios, and 6 model architectures; (iv) key insights into the factors that govern backdoor effectiveness and failure modes in LLMs; and (v) a defense toolkit encompassing 7 representative mitigation techniques. Our code and datasets are available at https://github.com/bboylyg/BackdoorLLM. We will continuously incorporate emerging attack and defense methodologies to support the research in advancing the safety and reliability of LLMs.
Embedding models play a crucial role in Natural Language Processing (NLP) by creating text embeddings used in various tasks such as information retrieval and assessing semantic text similarity. This paper focuses on research related to embedding models in the Russian language. It introduces a new Russian-focused embedding model called ru-en-RoSBERTa and the ruMTEB benchmark, the Russian version extending the Massive Text Embedding Benchmark (MTEB). Our benchmark includes seven categories of tasks, such as semantic textual similarity, text classification, reranking, and retrieval.The research also assesses a representative set of Russian and multilingual models on the proposed benchmark. The findings indicate that the new model achieves results that are on par with state-of-the-art models in Russian. We release the model ru-en-RoSBERTa, and the ruMTEB framework comes with open-source code, integration into the original framework and a public leaderboard.
In the era of large language models (LLMs), the task of ``System I''~-~the fast, unconscious, and intuitive tasks, e.g., sentiment analysis, text classification, etc., have been argued to be successfully solved. However, sarcasm, as a subtle linguistic phenomenon, often employs rhetorical devices like hyperbole and figuration to convey true sentiments and intentions, involving a higher level of abstraction than sentiment analysis. There is growing concern that the argument about LLMs' success may not be fully tenable when considering sarcasm understanding. To address this question, we select eleven SOTA LLMs and eight SOTA pre-trained language models (PLMs) and present comprehensive evaluations on six widely used benchmark datasets through different prompting approaches, i.e., zero-shot input/output (IO) prompting, few-shot IO prompting, chain of thought (CoT) prompting. Our results highlight three key findings: (1) current LLMs underperform supervised PLMs based sarcasm detection baselines across six sarcasm benchmarks. This suggests that significant efforts are still required to improve LLMs' understanding of human sarcasm. (2) GPT-4 consistently and significantly outperforms other LLMs across various prompting methods, with an average improvement of 14.0\%$\uparrow$. Claude 3 and ChatGPT demonstrate the next best performance after GPT-4. (3) Few-shot IO prompting method outperforms the other two methods: zero-shot IO and few-shot CoT. The reason is that sarcasm detection, being a holistic, intuitive, and non-rational cognitive process, is argued not to adhere to step-by-step logical reasoning, making CoT less effective in understanding sarcasm compared to its effectiveness in mathematical reasoning tasks.
In Explainable AI (XAI), counterfactual explanations (CEs) are a well-studied method to communicate feature relevance through contrastive reasoning of "what if" to explain AI models' predictions. However, they only focus on important (i.e., relevant) features and largely disregard less important (i.e., irrelevant) ones. Such irrelevant features can be crucial in many applications, especially when users need to ensure that an AI model's decisions are not affected or biased against specific attributes such as gender, race, religion, or political affiliation. To address this gap, the concept of alterfactual explanations (AEs) has been proposed. AEs explore an alternative reality of "no matter what", where irrelevant features are substituted with alternative features (e.g., "republicans" -> "democrats") within the same attribute (e.g., "politics") while maintaining a similar prediction output. This serves to validate whether AI model predictions are influenced by the specified attributes. Despite the promise of AEs, there is a lack of computational approaches to systematically generate them, particularly in the text domain, where creating AEs for AI text classifiers presents unique challenges. This paper addresses this challenge by formulating AE generation as an optimization problem and introducing MoMatterXAI, a novel algorithm that generates AEs for text classification tasks. Our approach achieves high fidelity of up to 95% while preserving context similarity of over 90% across multiple models and datasets. A human study further validates the effectiveness of AEs in explaining AI text classifiers to end users. All codes will be publicly available.
Evaluating the output of generative large language models (LLMs) is challenging and difficult to scale. Many evaluations of LLMs focus on tasks such as single-choice question-answering or text classification. These tasks are not suitable for assessing open-ended question-answering capabilities, which are critical in domains where expertise is required. One such domain is health, where misleading or incorrect answers can have a negative impact on a user's well-being. Using human experts to evaluate the quality of LLM answers is generally considered the gold standard, but expert annotation is costly and slow. We present a method for evaluating LLM answers that uses ranking models trained on annotated document collections as a substitute for explicit relevance judgements and apply it to the CLEF 2021 eHealth dataset. In a user study, our method correlates with the preferences of a human expert (Kendall's $\tau=0.64$). It is also consistent with previous findings in that the quality of generated answers improves with the size of the model and more sophisticated prompting strategies.