General Purpose Emotion Lexicons (GPELs) that associate words with emotion categories remain a valuable resource for emotion analysis of text. However the static and formal nature of their vocabularies make them inadequate for extracting effective features for document representation, in domains that are inherently dynamic in nature (e.g. Social Media). This calls for lexicons that are not only adaptive to the lexical variations in a domain but also provide finer-grained quantitative estimates to accurately capture word-emotion associations. In this paper we extend prior work on domain specific emotion lexicon (DSEL) generation and apply it for emotion feature extraction. We demonstrate how our generative unigram mixture model (UMM) based DSEL learnt by harnessing labelled (blogs, news headlines and incident reports) and weakly-labelled (tweets) emotion text can be used to extract effctive features for emotion classification. Our results confirm that the features derived using the proposed lexicon outperform those from state-of-the-art lexicons learnt using supervised Latent Dirichlet Allocation (sLDA) and Point-Wise Mutual Information (PMI). Further the proposed lexicon features also outperform state-of-the-art features derived using a combination of n-grams, part-of-speech information and sentiment lexicons
General-purpose emotion lexicons (GPELs) that associate words with emotion categories remain a valuable resource for emotion detection. However, the static and formal nature of their vocabularies make them an inadequate resource for detecting emotions in domains that are inherently dynamic in nature. This calls for lexicons that are not only adaptive to the lexical variations in a domain but which also provide finer-grained quantitative estimates to accurately capture word-emotion associations. In this article, the authors demonstrate how to harness labeled emotion text (such as blogs and news headlines) and weakly labeled emotion text (such as tweets) to learn a word-emotion association lexicon by jointly modeling emotionality and neutrality of words using a generative unigram mixture model (UMM). Empirical evaluation confirms that UMM generated emotion language models (topics) have significantly lower perplexity compared to those from state-of-the-art generative models like supervised Latent Dirichlet Allocation (sLDA). Further emotion detection tasks involving word-emotion classification and document-emotion ranking confirm that the UMM lexicon significantly out performs GPELs and also state-of-the-art domain specific lexicons.
The number of methods for identifying potential fall risk is growing as the rate of elderly fallers continues to rise in the UK. Assessments for identifying risk of falling are usually performed in hospitals and other laboratory environments, however these are costly and cause inconvenience for the subject and health services. Replacing these intrusive testing methods with a passive in-home monitoring solution would provide a less time-consuming and cheaper alternative. As sensors become more readily available, machine learning models can be applied to the large amount of data they produce. This can support activity recognition, falls detection, prediction and risk determination. In this review, the growing complexity of sensor data, the required analysis, and the machine learning techniques used to determine risk of falling are explored. The current research on using passive monitoring in the home is discussed, while the viability of active monitoring using vision-based and wearable sensors is considered. Methods of fall detection, prediction and risk determination are then compared.
The need to adhere to recommended physical activity guidelines for a variety of chronic disorders calls for high precision Human Activity Recognition (HAR) systems. In the SelfBACK system, HAR is used to monitor activity types and intensities to enable self-management of low back pain (LBP). HAR is typically modelled as a classification task where sensor data associated with activity labels are used to train a classifier to predict future occurrences of those activities. An important consideration in HAR is whether to use training data from a general population (subject-independent), or personalised training data from the target user (subject-dependent). Previous evaluations have shown that using personalised data results in more accurate predictions. However, from a practical perspective, collecting sufficient training data from the end user may not be feasible. This has made using subject-independent data by far the more common approach in commercial HAR systems. In this paper, we introduce a novel approach which uses nearest neighbour similarity to identify examples from a subject-independent training set that are most similar to sample data obtained from the target user and uses these examples to generate a personalised model for the user. This nearest neighbour sampling approach enables us to avoid much of the practical limitations associated with training a classifier exclusively with user data, while still achieving the benefit of personalisation. Evaluations show our approach to significantly out perform a general subject-independent model by up to 5%.
SelfBACK is an mHealth decision support system used by patients for the self-management of Lower Back Pain. It uses Human Activity Recognition from wearable sensors to monitor user activity in order to measure their adherence to prescribed physical activity plans. Different feature representation approaches have been proposed for Human Activity Recognition, including shallow, such as with hand-crafted time domain features and frequency transformation features; or, more recently, deep with Convolutional Neural Net approaches. The different approaches have produced mixed results in previous work and a clear winner has not been identified. This is especially the case for wrist mounted accelerometer sensors which are more susceptible to random noise compared to data from sensors mounted at other body locations e.g. thigh, waist or lower back. In this paper, we compare 7 different feature representation approaches on accelerometer data collected from both the wrist and the thigh. In particular, we evaluate a Convolutional Neural Net hybrid approach that has been shown to be effective on image retrieval but not previously applied to Human Activity Recognition. Results show the hybrid approach is effective, producing the best results compared to both hand-crafted and frequency domain feature representations by a margin of over \(1.4\%\) on the wrist.