We have two main contributions in this work: 1. We explore the usage of a stacked denoising autoencoder, and a paragraph vector model to learn task-independent dense patient representations directly from clinical notes. We evaluate these representations by using them as features in multiple supervised setups, and compare their performance with those of sparse representations. 2. To understand and interpret the representations, we explore the best encoded features within the patient representations obtained from the autoencoder model. Further, we calculate the significance of the input features of the trained classifiers when we use these pretrained representations as input.
We present an unsupervised context-sensitive spelling correction method for clinical free-text that uses word and character n-gram embeddings. Our method generates misspelling replacement candidates and ranks them according to their semantic fit, by calculating a weighted cosine similarity between the vectorized representation of a candidate and the misspelling context. To tune the parameters of this model, we generate self-induced spelling error corpora. We perform our experiments for two languages. For English, we greatly outperform off-the-shelf spelling correction tools on a manually annotated MIMIC-III test set, and counter the frequency bias of a noisy channel model, showing that neural embeddings can be successfully exploited to improve upon the state-of-the-art. For Dutch, we also outperform an off-the-shelf spelling correction tool on manually annotated clinical records from the Antwerp University Hospital, but can offer no empirical evidence that our method counters the frequency bias of a noisy channel model in this case as well. However, both our context-sensitive model and our implementation of the noisy channel model obtain high scores on the test set, establishing a state-of-the-art for Dutch clinical spelling correction with the noisy channel model.
Clinical NLP has an immense potential in contributing to how clinical practice will be revolutionized by the advent of large scale processing of clinical records. However, this potential has remained largely untapped due to slow progress primarily caused by strict data access policies for researchers. In this paper, we discuss the concern for privacy and the measures it entails. We also suggest sources of less sensitive data. Finally, we draw attention to biases that can compromise the validity of empirical research and lead to socially harmful applications.
We present a dictionary-based approach to racism detection in Dutch social media comments, which were retrieved from two public Belgian social media sites likely to attract racist reactions. These comments were labeled as racist or non-racist by multiple annotators. For our approach, three discourse dictionaries were created: first, we created a dictionary by retrieving possibly racist and more neutral terms from the training data, and then augmenting these with more general words to remove some bias. A second dictionary was created through automatic expansion using a \texttt{word2vec} model trained on a large corpus of general Dutch text. Finally, a third dictionary was created by manually filtering out incorrect expansions. We trained multiple Support Vector Machines, using the distribution of words over the different categories in the dictionaries as features. The best-performing model used the manually cleaned dictionary and obtained an F-score of 0.46 for the racist class on a test set consisting of unseen Dutch comments, retrieved from the same sites used for the training set. The automated expansion of the dictionary only slightly boosted the model's performance, and this increase in performance was not statistically significant. The fact that the coverage of the expanded dictionaries did increase indicates that the words that were automatically added did occur in the corpus, but were not able to meaningfully impact performance. The dictionaries, code, and the procedure for requesting the corpus are available at: https://github.com/clips/hades
In this paper, we report a knowledge-based method for Word Sense Disambiguation in the domains of biomedical and clinical text. We combine word representations created on large corpora with a small number of definitions from the UMLS to create concept representations, which we then compare to representations of the context of ambiguous terms. Using no relational information, we obtain comparable performance to previous approaches on the MSH-WSD dataset, which is a well-known dataset in the biomedical domain. Additionally, our method is fast and easy to set up and extend to other domains. Supplementary materials, including source code, can be found at https: //github.com/clips/yarn
Word embeddings have recently seen a strong increase in interest as a result of strong performance gains on a variety of tasks. However, most of this research also underlined the importance of benchmark datasets, and the difficulty of constructing these for a variety of language-specific tasks. Still, many of the datasets used in these tasks could prove to be fruitful linguistic resources, allowing for unique observations into language use and variability. In this paper we demonstrate the performance of multiple types of embeddings, created with both count and prediction-based architectures on a variety of corpora, in two language-specific tasks: relation evaluation, and dialect identification. For the latter, we compare unsupervised methods with a traditional, hand-crafted dictionary. With this research, we provide the embeddings themselves, the relation evaluation task benchmark for use in further research, and demonstrate how the benchmarked embeddings prove a useful unsupervised linguistic resource, effectively used in a downstream task.
The goal of this paper is to investigate the connection between the performance gain that can be obtained by selftraining and the similarity between the corpora used in this approach. Self-training is a semi-supervised technique designed to increase the performance of machine learning algorithms by automatically classifying instances of a task and adding these as additional training material to the same classifier. In the context of language processing tasks, this training material is mostly an (annotated) corpus. Unfortunately self-training does not always lead to a performance increase and whether it will is largely unpredictable. We show that the similarity between corpora can be used to identify those setups for which self-training can be beneficial. We consider this research as a step in the process of developing a classifier that is able to adapt itself to each new test corpus that it is presented with.
We present a corpus-based analysis of the effects of age, gender and region of origin on the production of both "netspeak" or "chatspeak" features and regional speech features in Flemish Dutch posts that were collected from a Belgian online social network platform. The present study shows that combining quantitative and qualitative approaches is essential for understanding non-standard linguistic variation in a CMC corpus. It also presents a methodology that enables the systematic study of this variation by including all non-standard words in the corpus. The analyses resulted in a convincing illustration of the Adolescent Peak Principle. In addition, our approach revealed an intriguing correlation between the use of regional speech features and chatspeak features.
We apply rule induction, classifier combination and meta-learning (stacked classifiers) to the problem of bootstrapping high accuracy automatic annotation of corpora with pronunciation information. The task we address in this paper consists of generating phonemic representations reflecting the Flemish and Dutch pronunciations of a word on the basis of its orthographic representation (which in turn is based on the actual speech recordings). We compare several possible approaches to achieve the text-to-pronunciation mapping task: memory-based learning, transformation-based learning, rule induction, maximum entropy modeling, combination of classifiers in stacked learning, and stacking of meta-learners. We are interested both in optimal accuracy and in obtaining insight into the linguistic regularities involved. As far as accuracy is concerned, an already high accuracy level (93% for Celex and 86% for Fonilex at word level) for single classifiers is boosted significantly with additional error reductions of 31% and 38% respectively using combination of classifiers, and a further 5% using combination of meta-learners, bringing overall word level accuracy to 96% for the Dutch variant and 92% for the Flemish variant. We also show that the application of machine learning methods indeed leads to increased insight into the linguistic regularities determining the variation between the two pronunciation variants studied.
We use seven machine learning algorithms for one task: identifying base noun phrases. The results have been processed by different system combination methods and all of these outperformed the best individual result. We have applied the seven learners with the best combinator, a majority vote of the top five systems, to a standard data set and managed to improve the best published result for this data set.