In the health-related field, drug non-compliance situations happen when patients do not follow their prescriptions and do actions which lead to potentially harmful situations. Although such situations are dangerous, patients usually do not report them to their physicians. Hence, it is necessary to study other sources of information. We propose to study online health fora with information retrieval methods in order to identify messages that contain drug non-compliance information. Exploitation of information retrieval methods permits to detect non-compliance messages with up to 0.529 F-measure, compared to 0.824 F-measure reached with supervized machine learning methods. For some fine-grained categories and on new data, it shows up to 0.70 Precision.
This contribution analyses the role of sentential context in reading intercomprehension both from an information-theoretic and an error-analytical perspective. The assumption is that not only cross-lingual similarity can influence the successful word disambiguation in an unknown but related foreign language, but also that predictability in context contributes to better intelligibility of the target items. Experimental data were gathered for 149 Polish sentences [1] with highly predictable target words in sentence final position presented to Czech readers in a web-based cloze translation task. Psycholinguistic research showed that predictably of words in context correlates with cognitive effort to process the information provided by the word and its surprisal [3]. Our hypothesis is that intelligibility of highly predictable words in sentential context of a related language also correlates with surprisal values obtained from statistical trigram language models. In order to establish a baseline, the individual words were also presented to Czech readers in a context-free translation experiment [4]. For the majority of the target words, an increase in correct translations is observable in context, as opposed to the results obtained without context. The overall correlations with surprisal are low, the highest being the joint surprisal of the Polish stimulus sentence. The error-analysis shows systematic patterns that are at least equally important intercomprehension factors, such as linguistic distance or morphological mismatches.
Self Normalizing Neural Networks(SNN) proposed on Feed Forward Neural Networks(FNN) outperform regular FNN architectures in various machine learning tasks. Particularly in the domain of Computer Vision, the activation function Scaled Exponential Linear Units (SELU) proposed for SNNs, perform better than other non linear activations such as ReLU. The goal of SNN is to produce a normalized output for a normalized input. Established neural network architectures like feed forward networks and Convolutional Neural Networks(CNN) lack the intrinsic nature of normalizing outputs. Hence, requiring additional layers such as Batch Normalization. Despite the success of SNNs, their characteristic features on other network architectures like CNN haven't been explored, especially in the domain of Natural Language Processing. In this paper we aim to show the effectiveness of proposed, Self Normalizing Convolutional Neural Networks(SCNN) on text classification. We analyze their performance with the standard CNN architecture used on several text classification datasets. Our experiments demonstrate that SCNN achieves comparable results to standard CNN model with significantly fewer parameters. Furthermore it also outperforms CNN with equal number of parameters.
We explore the abilities of character recurrent neural network (char-RNN) for hashtag segmentation. Our approach to the task is the following: we generate synthetic training dataset according to frequent n -grams that satisfy predefined morpho-syntactic patterns to avoid any manual annotation. The active learning strategy limits the training dataset and selects informative training subset. The approach does not require any language-specific settings and is compared for two languages, which differ in inflection degree.
The automatic summarization of multimedia sources is an important task that facilitates the understanding of an individual by condensing the source while maintaining relevant information. In this paper we focus on audio summarization based on audio features and the probability of distribution divergence. Our method, based on an extractive summarization approach, aims to select the most relevant segments until a time threshold is reached. It takes into account the segment’s length, position and informativeness value. Informativeness of each segment is obtained by mapping a set of audio features issued from its Mel-frequency Cepstral Coefficients and their corresponding Jensen-Shannon divergence score. Results over a multi-evaluator scheme shows that our approach provides understandable and informative summaries.
