Weakly supervised text classification (WSTC), also called zero-shot or dataless text classification, has attracted increasing attention due to its applicability in classifying a mass of texts within the dynamic and open Web environment, since it requires only a limited set of seed words (label names) for each category instead of labeled data. With the help of recently popular prompting Pre-trained Language Models (PLMs), many studies leveraged manually crafted and/or automatically identified verbalizers to estimate the likelihood of categories, but they failed to differentiate the effects of these category-indicative words, let alone capture their correlations and realize adaptive adjustments according to the unlabeled corpus. In this paper, in order to let the PLM effectively understand each category, we at first propose a novel form of rule-based knowledge using logical expressions to characterize the meanings of categories. Then, we develop a prompting PLM-based approach named RulePrompt for the WSTC task, consisting of a rule mining module and a rule-enhanced pseudo label generation module, plus a self-supervised fine-tuning module to make the PLM align with this task. Within this framework, the inaccurate pseudo labels assigned to texts and the imprecise logical rules associated with categories mutually enhance each other in an alternative manner. That establishes a self-iterative closed loop of knowledge (rule) acquisition and utilization, with seed words serving as the starting point. Extensive experiments validate the effectiveness and robustness of our approach, which markedly outperforms state-of-the-art weakly supervised methods. What is more, our approach yields interpretable category rules, proving its advantage in disambiguating easily-confused categories.
Attention mechanisms have revolutionized natural language processing. Combining them with quantum computing aims to further advance this technology. This paper introduces a novel Quantum Mixed-State Self-Attention Network (QMSAN) for natural language processing tasks. Our model leverages quantum computing principles to enhance the effectiveness of self-attention mechanisms. QMSAN uses a quantum attention mechanism based on mixed state, allowing for direct similarity estimation between queries and keys in the quantum domain. This approach leads to more effective attention coefficient calculations. We also propose an innovative quantum positional encoding scheme, implemented through fixed quantum gates within the circuit, improving the model's ability to capture sequence information without additional qubit resources. In numerical experiments of text classification tasks on public datasets, QMSAN outperforms Quantum Self-Attention Neural Network (QSANN). Furthermore, we demonstrate QMSAN's robustness in different quantum noise environments, highlighting its potential for near-term quantum devices.
Adaptive methods are extremely popular in machine learning as they make learning rate tuning less expensive. This paper introduces a novel optimization algorithm named KATE, which presents a scale-invariant adaptation of the well-known AdaGrad algorithm. We prove the scale-invariance of KATE for the case of Generalized Linear Models. Moreover, for general smooth non-convex problems, we establish a convergence rate of $O \left(\frac{\log T}{\sqrt{T}} \right)$ for KATE, matching the best-known ones for AdaGrad and Adam. We also compare KATE to other state-of-the-art adaptive algorithms Adam and AdaGrad in numerical experiments with different problems, including complex machine learning tasks like image classification and text classification on real data. The results indicate that KATE consistently outperforms AdaGrad and matches/surpasses the performance of Adam in all considered scenarios.
Text classification is the task of assigning a document to a predefined class. However, it is expensive to acquire enough labeled documents or to label them. In this paper, we study the regularization methods' effects on various classification models when only a few labeled data are available. We compare a simple word embedding-based model, which is simple but effective, with complex models (CNN and BiLSTM). In supervised learning, adversarial training can further regularize the model. When an unlabeled dataset is available, we can regularize the model using semi-supervised learning methods such as the Pi model and virtual adversarial training. We evaluate the regularization effects on four text classification datasets (AG news, DBpedia, Yahoo! Answers, Yelp Polarity), using only 0.1% to 0.5% of the original labeled training documents. The simple model performs relatively well in fully supervised learning, but with the help of adversarial training and semi-supervised learning, both simple and complex models can be regularized, showing better results for complex models. Although the simple model is robust to overfitting, a complex model with well-designed prior beliefs can be also robust to overfitting.
Multi-domain text classification (MDTC) endeavors to harness available resources from correlated domains to enhance the classification accuracy of the target domain. Presently, most MDTC approaches that embrace adversarial training and the shared-private paradigm exhibit cutting-edge performance. Unfortunately, these methods face a non-negligible challenge: the absence of theoretical guarantees in the design of MDTC algorithms. The dearth of theoretical underpinning poses a substantial impediment to the advancement of MDTC algorithms. To tackle this problem, we first provide a theoretical analysis of MDTC by decomposing the MDTC task into multiple domain adaptation tasks. We incorporate the margin discrepancy as the measure of domain divergence and establish a new generalization bound based on Rademacher complexity. Subsequently, we propose a margin discrepancy-based adversarial training (MDAT) approach for MDTC, in accordance with our theoretical analysis. To validate the efficacy of the proposed MDAT method, we conduct empirical studies on two MDTC benchmarks. The experimental results demonstrate that our MDAT approach surpasses state-of-the-art baselines on both datasets.
Hierarchical text classification aims to categorize each document into a set of classes in a label taxonomy. Most earlier works focus on fully or semi-supervised methods that require a large amount of human annotated data which is costly and time-consuming to acquire. To alleviate human efforts, in this paper, we work on hierarchical text classification with the minimal amount of supervision: using the sole class name of each node as the only supervision. Recently, large language models (LLM) show competitive performance on various tasks through zero-shot prompting, but this method performs poorly in the hierarchical setting, because it is ineffective to include the large and structured label space in a prompt. On the other hand, previous weakly-supervised hierarchical text classification methods only utilize the raw taxonomy skeleton and ignore the rich information hidden in the text corpus that can serve as additional class-indicative features. To tackle the above challenges, we propose TELEClass, Taxonomy Enrichment and LLM-Enhanced weakly-supervised hierarchical text Classification, which (1) automatically enriches the label taxonomy with class-indicative terms to facilitate classifier training and (2) utilizes LLMs for both data annotation and creation tailored for the hierarchical label space. Experiments show that TELEClass can outperform previous weakly-supervised methods and LLM-based zero-shot prompting methods on two public datasets.
Traditional dataset retrieval systems rely on metadata for indexing, rather than on the underlying data values. However, high-quality metadata creation and enrichment often require manual annotations, which is a labour-intensive and challenging process to automate. In this study, we propose a method to support metadata enrichment using topic annotations generated by three Large Language Models (LLMs): ChatGPT-3.5, GoogleBard, and GoogleGemini. Our analysis focuses on classifying column headers based on domain-specific topics from the Consortium of European Social Science Data Archives (CESSDA), a Linked Data controlled vocabulary. Our approach operates in a zero-shot setting, integrating the controlled topic vocabulary directly within the input prompt. This integration serves as a Large Context Windows approach, with the aim of improving the results of the topic classification task. We evaluated the performance of the LLMs in terms of internal consistency, inter-machine alignment, and agreement with human classification. Additionally, we investigate the impact of contextual information (i.e., dataset description) on the classification outcomes. Our findings suggest that ChatGPT and GoogleGemini outperform GoogleBard in terms of internal consistency as well as LLM-human-agreement. Interestingly, we found that contextual information had no significant impact on LLM performance. This work proposes a novel approach that leverages LLMs for topic classification of column headers using a controlled vocabulary, presenting a practical application of LLMs and Large Context Windows within the Semantic Web domain. This approach has the potential to facilitate automated metadata enrichment, thereby enhancing dataset retrieval and the Findability, Accessibility, Interoperability, and Reusability (FAIR) of research data on the Web.
Hierarchical text classification (HTC) is a challenging subtask of multi-label classification due to its complex taxonomic structure. Nearly all recent HTC works focus on how the labels are structured but ignore the sub-structure of ground-truth labels according to each input text which contains fruitful label co-occurrence information. In this work, we introduce this local hierarchy with an adversarial framework. We propose a HiAdv framework that can fit in nearly all HTC models and optimize them with the local hierarchy as auxiliary information. We test on two typical HTC models and find that HiAdv is effective in all scenarios and is adept at dealing with complex taxonomic hierarchies. Further experiments demonstrate that the promotion of our framework indeed comes from the local hierarchy and the local hierarchy is beneficial for rare classes which have insufficient training data.
Language models (LMs) are statistical models trained to assign probability to human-generated text. As such, it is reasonable to question whether they approximate linguistic variability exhibited by humans well. This form of statistical assessment is difficult to perform at the passage level, for it requires acceptability judgements (i.e., human evaluation) or a robust automated proxy (which is non-trivial). At the word level, however, given some context, samples from an LM can be assessed via exact matching against a prerecorded dataset of alternative single-word continuations of the available context. We exploit this fact and evaluate the LM's ability to reproduce variability that humans (in particular, a population of English speakers) exhibit in the 'next word prediction' task. This can be seen as assessing a form of calibration, which, in the context of text classification, Baan et al. (2022) termed calibration to human uncertainty. We assess GPT2, BLOOM and ChatGPT and find that they exhibit fairly low calibration to human uncertainty. We also verify the failure of expected calibration error (ECE) to reflect this, and as such, advise the community against relying on it in this setting.
As sufficient data are not always publically accessible for model training, researchers exploit limited data with advanced learning algorithms or expand the dataset via data augmentation (DA). Conducting DA in private domain requires private protection approaches (i.e. anonymization and perturbation), but those methods cannot provide protection guarantees. Differential privacy (DP) learning methods theoretically bound the protection but are not skilled at generating pseudo text samples with large models. In this paper, we transfer DP-based pseudo sample generation task to DP-based generated samples discrimination task, where we propose a DP-based DA method with a LLM and a DP-based discriminator for text classification on private domains. We construct a knowledge distillation model as the DP-based discriminator: teacher models, accessing private data, teaches students how to select private samples with calibrated noise to achieve DP. To constrain the distribution of DA's generation, we propose a DP-based tutor that models the noised private distribution and controls samples' generation with a low privacy cost. We theoretically analyze our model's privacy protection and empirically verify our model.
Fine-tuning Large Language Models (LLMs) is now a common approach for text classification in a wide range of applications. When labeled documents are scarce, active learning helps save annotation efforts but requires retraining of massive models on each acquisition iteration. We drastically expedite this process by using pretrained representations of LLMs within the active learning loop and, once the desired amount of labeled data is acquired, fine-tuning that or even a different pretrained LLM on this labeled data to achieve the best performance. As verified on common text classification benchmarks with pretrained BERT and RoBERTa as the backbone, our strategy yields similar performance to fine-tuning all the way through the active learning loop but is orders of magnitude less computationally expensive. The data acquired with our procedure generalizes across pretrained networks, allowing flexibility in choosing the final model or updating it as newer versions get released.
The availability of large annotated data can be a critical bottleneck in training machine learning algorithms successfully, especially when applied to diverse domains. Weak supervision offers a promising alternative by accelerating the creation of labeled training data using domain-specific rules. However, it requires users to write a diverse set of high-quality rules to assign labels to the unlabeled data. Automatic Rule Induction (ARI) approaches circumvent this problem by automatically creating rules from features on a small labeled set and filtering a final set of rules from them. In the ARI approach, the crucial step is to filter out a set of a high-quality useful subset of rules from the large set of automatically created rules. In this paper, we propose an algorithm (Filtering of Automatically Induced Rules) to filter rules from a large number of automatically induced rules using submodular objective functions that account for the collective precision, coverage, and conflicts of the rule set. We experiment with three ARI approaches and five text classification datasets to validate the superior performance of our algorithm with respect to several semi-supervised label aggregation approaches. Further, we show that achieves statistically significant results in comparison to existing rule-filtering approaches.
Textual data augmentation (DA) is a prolific field of study where novel techniques to create artificial data are regularly proposed, and that has demonstrated great efficiency on small data settings, at least for text classification tasks. In this paper, we challenge those results, showing that classical data augmentation (which modify sentences) is simply a way of performing better fine-tuning, and that spending more time doing so before applying data augmentation negates its effect. This is a significant contribution as it answers several questions that were left open in recent years, namely~: which DA technique performs best (all of them as long as they generate data close enough to the training set, as to not impair training) and why did DA show positive results (facilitates training of network). We further show that zero- and few-shot DA via conversational agents such as ChatGPT or LLama2 can increase performances, confirming that this form of data augmentation is preferable to classical methods.
Classifying text is a method for categorizing documents into pre-established groups. Text documents must be prepared and represented in a way that is appropriate for the algorithms used for data mining prior to classification. As a result, a number of term weighting strategies have been created in the literature to enhance text categorization algorithms' functionality. This study compares the effects of Binary and Term frequency weighting feature methodologies on the text's classification method when stop words are eliminated once and when they are not. In recognition of assessing the effects of prior weighting of features approaches on classification results in terms of accuracy, recall, precision, and F-measure values, we used an Arabic data set made up of 322 documents divided into six main topics (agriculture, economy, health, politics, science, and sport), each of which contains 50 documents, with the exception of the health category, which contains 61 documents. The results demonstrate that for all metrics, the term frequency feature weighting approach with stop word removal outperforms the binary approach, while for accuracy, recall, and F-Measure, the binary approach outperforms the TF approach without stop word removal. However, for precision, the two approaches produce results that are very similar. Additionally, it is clear from the data that, using the same phrase weighting approach, stop word removing increases classification accuracy.
In machine learning, ensembles are important tools for improving the model performance. In natural language processing specifically, ensembles boost the performance of a method due to multiple large models available in open source. However, existing approaches mostly rely on simple averaging of predictions by ensembles with equal weights for each model, ignoring differences in the quality and conformity of models. We propose to estimate weights for ensembles of NLP models using not only knowledge of their individual performance but also their similarity to each other. By adopting distance measures based on Topological Data Analysis (TDA), we improve our ensemble. The quality improves for both text classification accuracy and relevant uncertainty estimation.
SemEval-2024 Task 8 is focused on multigenerator, multidomain, and multilingual black-box machine-generated text detection. Such a detection is important for preventing a potential misuse of large language models (LLMs), the newest of which are very capable in generating multilingual human-like texts. We have coped with this task in multiple ways, utilizing language identification and parameter-efficient fine-tuning of smaller LLMs for text classification. We have further used the per-language classification-threshold calibration to uniquely combine fine-tuned models predictions with statistical detection metrics to improve generalization of the system detection performance. Our submitted method achieved competitive results, ranking at the fourth place, just under 1 percentage point behind the winner.
As supervised fine-tuning of pre-trained models within NLP applications increases in popularity, larger corpora of annotated data are required, especially with increasing parameter counts in large language models. Active learning, which attempts to mine and annotate unlabeled instances to improve model performance maximally fast, is a common choice for reducing the annotation cost; however, most methods typically ignore class imbalance and either assume access to initial annotated data or require multiple rounds of active learning selection before improving rare classes. We present STENCIL, which utilizes a set of text exemplars and the recently proposed submodular mutual information to select a set of weakly labeled rare-class instances that are then strongly labeled by an annotator. We show that STENCIL improves overall accuracy by $10\%-18\%$ and rare-class F-1 score by $17\%-40\%$ on multiple text classification datasets over common active learning methods within the class-imbalanced cold-start setting.
Besides natural language processing, transformers exhibit extraordinary performance in solving broader applications, including scientific computing and computer vision. Previous works try to explain this from the expressive power and capability perspectives that standard transformers are capable of performing some algorithms. To empower transformers with algorithmic capabilities and motivated by the recently proposed looped transformer, we design a novel transformer framework, dubbed Algorithm Transformer (abbreviated as AlgoFormer). We provide an insight that efficient transformer architectures can be designed by leveraging prior knowledge of tasks and the underlying structure of potential algorithms. Compared with the standard transformer and vanilla looped transformer, the proposed AlgoFormer can perform efficiently in algorithm representation in some specific tasks. In particular, inspired by the structure of human-designed learning algorithms, our transformer framework consists of a pre-transformer that is responsible for task preprocessing, a looped transformer for iterative optimization algorithms, and a post-transformer for producing the desired results after post-processing. We provide theoretical evidence of the expressive power of the AlgoFormer in solving some challenging problems, mirroring human-designed algorithms. Furthermore, some theoretical and empirical results are presented to show that the designed transformer has the potential to perform algorithm representation and learning. Experimental results demonstrate the empirical superiority of the proposed transformer in that it outperforms the standard transformer and vanilla looped transformer in some specific tasks. An extensive experiment on real language tasks (e.g., neural machine translation of German and English, and text classification) further validates the expressiveness and effectiveness of AlgoFormer.
Phishing email is a serious cyber threat that tries to deceive users by sending false emails with the intention of stealing confidential information or causing financial harm. Attackers, often posing as trustworthy entities, exploit technological advancements and sophistication to make detection and prevention of phishing more challenging. Despite extensive academic research, phishing detection remains an ongoing and formidable challenge in the cybersecurity landscape. Large Language Models (LLMs) and Masked Language Models (MLMs) possess immense potential to offer innovative solutions to address long-standing challenges. In this research paper, we present an optimized, fine-tuned transformer-based DistilBERT model designed for the detection of phishing emails. In the detection process, we work with a phishing email dataset and utilize the preprocessing techniques to clean and solve the imbalance class issues. Through our experiments, we found that our model effectively achieves high accuracy, demonstrating its capability to perform well. Finally, we demonstrate our fine-tuned model using Explainable-AI (XAI) techniques such as Local Interpretable Model-Agnostic Explanations (LIME) and Transformer Interpret to explain how our model makes predictions in the context of text classification for phishing emails.
When solving NLP tasks with limited labelled data, researchers typically either use a general large language model without further update, or use a small number of labelled samples to tune a specialised smaller model. In this work, we answer an important question -- how many labelled samples are required for the specialised small models to outperform general large models, while taking the performance variance into consideration. By observing the behaviour of fine-tuning, instruction-tuning, prompting and in-context learning on 8 language models, we identify such performance break-even points across 8 representative text classification tasks of varying characteristics. We show that the specialised models often need only few samples (on average $100$) to be on par or better than the general ones. At the same time, the number of required labels strongly depends on the dataset or task characteristics, with fine-tuning on binary datasets requiring significantly more samples. When performance variance is taken into consideration, the number of required labels increases on average by $100 - 200\%$. Finally, larger models do not consistently lead to better performance and lower variance, with 4-bit quantisation having negligible impact.