An accurate and substantial dataset is essential for training a reliable and well-performing model. However, even manually annotated datasets contain label errors, not to mention automatically labeled ones. Previous methods for label denoising have primarily focused on detecting outliers and their permanent removal - a process that is likely to over- or underfilter the dataset. In this work, we propose AGRA: a new method for learning with noisy labels by using Adaptive GRAdient-based outlier removal. Instead of cleaning the dataset prior to model training, the dataset is dynamically adjusted during the training process. By comparing the aggregated gradient of a batch of samples and an individual example gradient, our method dynamically decides whether a corresponding example is helpful for the model at this point or is counter-productive and should be left out for the current update. Extensive evaluation on several datasets demonstrates AGRA's effectiveness, while a comprehensive results analysis supports our initial hypothesis: permanent hard outlier removal is not always what model benefits the most from.
Self-supervised knowledge-graph completion (KGC) relies on estimating a scoring model over (entity, relation, entity)-tuples, for example, by embedding an initial knowledge graph. Prediction quality can be improved by calibrating the scoring model, typically by adjusting the prediction thresholds using manually annotated examples. In this paper, we attempt for the first time cold-start calibration for KGC, where no annotated examples exist initially for calibration, and only a limited number of tuples can be selected for annotation. Our new method ACTC finds good per-relation thresholds efficiently based on a limited set of annotated tuples. Additionally to a few annotated tuples, ACTC also leverages unlabeled tuples by estimating their correctness with Logistic Regression or Gaussian Process classifiers. We also experiment with different methods for selecting candidate tuples for annotation: density-based and random selection. Experiments with five scoring models and an oracle annotator show an improvement of 7% points when using ACTC in the challenging setting with an annotation budget of only 10 tuples, and an average improvement of 4% points over different budgets.
A way to overcome expensive and time-consuming manual data labeling is weak supervision - automatic annotation of data samples via a predefined set of labeling functions (LFs), rule-based mechanisms that generate potentially erroneous labels. In this work, we investigate noise reduction techniques for weak supervision based on the principle of k-fold cross-validation. In particular, we extend two frameworks for detecting the erroneous samples in manually annotated data to the weakly supervised setting. Our methods profit from leveraging the information about matching LFs and detect noisy samples more accurately. We also introduce a new algorithm for denoising the weakly annotated data called ULF, that refines the allocation of LFs to classes by estimating the reliable LFs-to-classes joint matrix. Evaluation on several datasets shows that ULF successfully improves weakly supervised learning without using any manually labeled data.
Strategies for improving the training and prediction quality of weakly supervised machine learning models vary in how much they are tailored to a specific task or integrated with a specific model architecture. In this work, we propose a software framework Knodle that treats weak data annotations, deep learning models, and methods for improving weakly supervised training as separate, modular components. The standardized interfaces between these independent parts account for data- and model-agnostic weak supervision method development, but still allow the training process to access fine-grained information such as data set characteristics, matches of heuristic rules, as well as elements of the deep learning model ultimately used for prediction. Hence, our framework can encompass a wide range of training methods for improving weak supervision, ranging from methods that only look at correlations of rules and output classes (independently of the machine learning model trained with the resulting labels), to those that harness the interplay of neural networks and weakly labeled data. We illustrate the benchmarking potential of the framework with a performance comparison of several reference implementations on a selection of datasets that are already available in Knodle.