Active learning is an iterative labeling process that is used to obtain a small labeled subset, despite the absence of labeled data, thereby enabling to train a model for supervised tasks such as text classification.While active learning has made considerable progress in recent years due to improvements provided by pre-trained language models, there is untapped potential in the often neglected unlabeled portion of the data, although it is available in considerably larger quantities than the usually small set of labeled data. In this work, we investigate how self-training, a semi-supervised approach that uses a model to obtain pseudo-labels for unlabeled data, can be used to improve the efficiency of active learning for text classification. Building on a comprehensive reproduction of four previous self-training approaches, some of which are evaluated for the first time in the context of active learning or natural language processing, we introduce HAST, a new and effective self-training strategy, which is evaluated on four text classification benchmarks. Our results show that it outperforms the reproduced self-training approaches and reaches classification results comparable to previous experiments for three out of four datasets, using as little as 25% of the data. The code is publicly available at https://github.com/chschroeder/self-training-for-sample-efficient-active-learning.

A trigger warning is used to warn people about potentially disturbing content. We introduce trigger warning assignment as a multi-label classification task, create the Webis Trigger Warning Corpus 2022, and with it the first dataset of 1 million fanfiction works from Archive of our Own with up to 36 different warnings per document. To provide a reliable catalog of trigger warnings, we organized 41 million of free-form tags assigned by fanfiction authors into the first comprehensive taxonomy of trigger warnings by mapping them to the 36 institutionally recommended warnings. To determine the best operationalization of trigger warnings, we explore state-of-the-art multi-label models, examining the trade-off between assigning coarse- and fine-grained warnings, open- and closed-set classification, document length, and label confidence. Our models achieve micro-F1 scores of about 0.5, which reveals the difficulty of the task. Tailored representations, long input sequences, and a higher recall on rare warnings would help.

We present the first dataset and evaluation results on a newly defined task: assigning trigger warnings. We introduce a labeled corpus of narrative fiction from Archive of Our Own (AO3), a popular fan fiction site, and define a document-level classification task to determine whether or not to assign a trigger warning to an English story. We focus on the most commonly assigned trigger type “violence’ using the warning labels provided by AO3 authors as ground-truth labels. We trained SVM, BERT, and Longfomer models on three datasets sampled from the corpus and achieve F1 scores between 0.8 and 0.9, indicating that assigning trigger warnings for violence is feasible.

We introduce small-text, an easy-to-use active learning library, which offers pool-based active learning for single- and multi-label text classification in Python. It features numerous pre-implemented state-of-the-art query strategies, including some that leverage the GPU. Standardized interfaces allow the combination of a variety of classifiers, query strategies, and stopping criteria, facilitating a quick mix and match, and enabling a rapid development of both active learning experiments and applications. With the objective of making various classifiers and query strategies accessible for active learning, small-text integrates several well-known machine learning libraries, namely scikit-learn, Pytorch, and Hugging Face transformers. The latter integrations are optionally installable extensions, so GPUs can be used but are not required. Using this new library, we investigate the performance of the recently published SetFit training paradigm, which we compare to vanilla transformer fine-tuning, finding that it matches the latter in classification accuracy while outperforming it in area under the curve. The library is available under the MIT License at https://github.com/webis-de/small-text, in version 1.3.0 at the time of writing.

Active learning is the iterative construction of a classification model through targeted labeling, enabling significant labeling cost savings. As most research on active learning has been carried out before transformer-based language models (“transformers”) became popular, despite its practical importance, comparably few papers have investigated how transformers can be combined with active learning to date. This can be attributed to the fact that using state-of-the-art query strategies for transformers induces a prohibitive runtime overhead, which effectively nullifies, or even outweighs the desired cost savings. For this reason, we revisit uncertainty-based query strategies, which had been largely outperformed before, but are particularly suited in the context of fine-tuning transformers. In an extensive evaluation, we connect transformers to experiments from previous research, assessing their performance on five widely used text classification benchmarks. For active learning with transformers, several other uncertainty-based approaches outperform the well-known prediction entropy query strategy, thereby challenging its status as most popular uncertainty baseline in active learning for text classification.

The “Web as corpus” paradigm opens opportunities for enhancing the current state of language resources for endangered and under-resourced languages. However, standard crawling strategies tend to overlook available resources of these languages in favor of already well-documented ones. Since 2016, the “Crawling Under-Resourced Languages” portal (CURL) has been contributing to bridging the gap between established crawling techniques and knowledge about relevant Web resources that is only available in the specific language communities. The aim of the CURL portal is to enlarge the amount of available text material for under-resourced languages thereby developing available datasets further and to use them as a basis for statistical evaluation and enrichment of already available resources. The application is currently provided and further developed as part of the thematic cluster “Non-Latin scripts and Under-resourced languages” in the German national research consortium Text+. In this context, its focus lies on the extraction of text material and statistical information for the data domain “Lexical resources”.

Open pit mines left many regions worldwide inhospitable or uninhabitable. Many sites are left behind in a hazardous or contaminated state, show remnants of waste, or have other restrictions imposed upon them, e.g., for the protection of human or nature. Such information has to be permanently managed in order to reuse those areas in the future. In this work we present and evaluate an automated workflow for supporting the post-mining management of former lignite open pit mines in the eastern part of Germany, where prior to any planned land reuse, aforementioned information has to be acquired to ensure the safety and validity of such an endeavor. Usually, this information is found in expert reports, either in the form of paper documents, or in the best case as digitized unstructured text—all of them in German language. However, due to the size and complexity of these documents, any inquiry is tedious and time-consuming, thereby slowing down or even obstructing the reuse of related areas. Since no training data is available, we employ active learning in order to perform multi-label sentence classification for two categories of restrictions and seven categories of topics. The final system integrates optical character recognition (OCR), active-learning-based text classification, and geographic information system visualization in order to effectively extract, query, and visualize this information for any area of interest. Active learning and text classification results are twofold: Whereas the restriction categories were reasonably accurate (>0.85 F1), the seven topic-oriented categories seemed to be complex even for human annotators and achieved mediocre evaluation scores (<0.70 F1).