Self-training from labeled features for sentiment analysis

Sentiment analysis concerns about automatically identifying sentiment or opinion expressed in a given piece of text. Most prior work either use prior lexical knowledge defined as sentiment polarity of words or view the task as a text classification problem and rely on labeled corpora to train a sentiment classifier. While lexicon-based approaches do not adapt well to different domains, corpus-based approaches require expensive manual annotation effort. In this paper, we propose a novel framework where an initial classifier is learned by incorporating prior information extracted from an existing sentiment lexicon with preferences on expectations of sentiment labels of those lexicon words being expressed using generalized expectation criteria. Documents classified with high confidence are then used as pseudo-labeled examples for automatical domain-specific feature acquisition. The word-class distributions of such self-learned features are estimated from the pseudo-labeled examples and are used to train another classifier by constraining the model's predictions on unlabeled instances. Experiments on both the movie-review data and the multi-domain sentiment dataset show that our approach attains comparable or better performance than existing weakly-supervised sentiment classification methods despite using no labeled documents.

Decentralized supply chain formation using max-sum loopy belief propagation

Supply chain formation is the process by which a set of producers within a network determine the subset of these producers able to form a chain to supply goods to one or more consumers at the lowest cost. This problem has been tackled in a number of ways, including auctions, negotiations, and argumentation-based approaches. In this paper we show how this problem can be cast as an optimization of a pairwise cost function. Optimizing this class of energy functions is NP-hard but efficient approximations to the global minimum can be obtained using loopy belief propagation (LBP). Here we detail a max-sum LBP-based approach to the supply chain formation problem, involving decentralized message-passing between supply chain participants. Our approach is evaluated against a well-known decentralized double-auction method and an optimal centralized technique, showing several improvements on the auction method: it obtains better solutions for most network instances which allow for competitive equilibrium (Competitive equilibrium in Walsh and Wellman is a set of producer costs which permits a Pareto optimal state in which agents in the allocation receive non-negative surplus and agents not in the allocation would acquire non-positive surplus by participating in the supply chain) while also optimally solving problems where no competitive equilibrium exists, for which the double-auction method frequently produces inefficient solutions. © 2012 Wiley Periodicals, Inc.

Magnetite-doped polydimethylsiloxane (PDMS) for phosphopeptide enrichment

Reversible phosphorylation plays a key role in numerous biological processes. Mass spectrometry-based approaches are commonly used to analyze protein phosphorylation, but such analysis is challenging, largely due to the low phosphorylation stoichiometry. Hence, a number of phosphopeptide enrichment strategies have been developed, including metal oxide affinity chromatography (MOAC). Here, we describe a new material for performing MOAC that employs a magnetite-doped polydimethylsiloxane (PDMS), that is suitable for the creation of microwell array and microfluidic systems to enable low volume, high throughput analysis. Incubation time and sample loading were explored and optimized and demonstrate that the embedded magnetite is able to enrich phosphopeptides. This substrate-based approach is rapid, straightforward and suitable for simultaneously performing multiple, low volume enrichments. © the Partner Organisations 2014.

Improving cancer services through patient involvement

This book provides a clear approach to establishing a user involvement system in a healthcare organisation and its potential impact on cancer services. Using a tool kit style approach, drawing on examples of successful past projects and case studies to provide evidence of good practice, it describes how to plan and implement different stages of user involvement, enabling organisations to draw on user experience and expertise to evaluate, develop and improve the quality of service that they provide. Members of regional cancer networks, multidisciplinary cancer care teams, and all those involved in the NHS cancer services will find this toolkit interesting reading.

A Method of Context-sensitive Help Generation Using a Task Project

The article presents a new method to automatic generation of help in software. Help generation is realized in the framework of the tool for development and automatic generation of user interfaces based on ontologies. The principal features of the approach are: support for context-sensitive help, automatic generation of help using a task project and an expandable system of help generation.

"Comparing the incomparable":hospital costing and the art of medicine in post-war Britain

This paper examines the relationship between medical and hospital accounting discourses during the two decades after the 1946 National Health Service (NHS) Act for England and Wales. It argues that the departmental costing system introduced into the NHS in 1957 was concerned with the administrative aspects of hospital costliness as contemporary hospital accountants suggested that the perceived incomparability, immeasurability and uncontrollability of medical practice precluded the application of cost accounting to the clinical functions of hospitals. The paper links these suggestions to medical discourses which portrayed the practice of medicine as an intuitive and experience-based art and argues that post-war conceptions of clinical medicine represented this domain in a manner that was neither susceptible to the calculations of cost accountants nor to calculating and normalising intervention more generally. The paper concludes by suggesting that a closer engagement with medical discourses may enhance our understanding of historical as well as present day attempts to make medicine calculable.

Contextual semantics for sentiment analysis of Twitter

Sentiment analysis on Twitter has attracted much attention recently due to its wide applications in both, commercial and public sectors. In this paper we present SentiCircles, a lexicon-based approach for sentiment analysis on Twitter. Different from typical lexicon-based approaches, which offer a fixed and static prior sentiment polarities of words regardless of their context, SentiCircles takes into account the co-occurrence patterns of words in different contexts in tweets to capture their semantics and update their pre-assigned strength and polarity in sentiment lexicons accordingly. Our approach allows for the detection of sentiment at both entity-level and tweet-level. We evaluate our proposed approach on three Twitter datasets using three different sentiment lexicons to derive word prior sentiments. Results show that our approach significantly outperforms the baselines in accuracy and F-measure for entity-level subjectivity (neutral vs. polar) and polarity (positive vs. negative) detections. For tweet-level sentiment detection, our approach performs better than the state-of-the-art SentiStrength by 4-5% in accuracy in two datasets, but falls marginally behind by 1% in F-measure in the third dataset.

Externalizing, internalizing and fostering commitment:the case of born-global firms in emerging economies

This paper examines the HR practices of mature born-global firms from 29 emerging economies. Through an examination of large scale survey data the paper questions the extent to which firm size impacts the employment of temporary workers, the employment of skilled workers and the extent of employee training. Findings suggest that as firm size increases the use of temporary workers decreases, the number of skilled workers increases and the number of employees receiving training also increases. The paper highlights how born-global firms are able to shift away from externalized, market-based approaches towards more internalized, commitment-based approaches in order to survive, adapt and grow.

SentiCircles for contextual and conceptual semantic sentiment analysis of Twitter

Lexicon-based approaches to Twitter sentiment analysis are gaining much popularity due to their simplicity, domain independence, and relatively good performance. These approaches rely on sentiment lexicons, where a collection of words are marked with fixed sentiment polarities. However, words' sentiment orientation (positive, neural, negative) and/or sentiment strengths could change depending on context and targeted entities. In this paper we present SentiCircle; a novel lexicon-based approach that takes into account the contextual and conceptual semantics of words when calculating their sentiment orientation and strength in Twitter. We evaluate our approach on three Twitter datasets using three different sentiment lexicons. Results show that our approach significantly outperforms two lexicon baselines. Results are competitive but inconclusive when comparing to state-of-art SentiStrength, and vary from one dataset to another. SentiCircle outperforms SentiStrength in accuracy on average, but falls marginally behind in F-measure. © 2014 Springer International Publishing.

Learning to research - researching to learn

Learning to Research Researching to Learn explores the integration of research into teaching and learning at all levels of higher education. The chapters draw on the long and ongoing debate about the teaching research nexus in universities. Although the vast majority of academics believe that there is an important and valuable link between teaching and research, the precise nature of this relationship continues to be contested. The book includes chapters that showcase innovative ways of learning to research; how research is integrated into coursework teaching; how students learn the processes of research, and how universities are preparing students to engage with the world. The chapters also showcase innovative ways of researching to learn, exploring how students learn through doing research, how they conceptualise the knowledge of their fields of study through the processes of doing research, and how students experiment and reflect on the results produced. These are the key issues addressed by this anthology, as it brings together analyses of the ways in which university teachers are developing research skills in their students, creating enquiry-based approaches to teaching, and engaging in education research themselves. The studies here explore the links between teaching, learning and research in a range of contexts, from pre-enrolment through to academic staff development, in Australia, the UK, the US, Singapore and Denmark. Through a rich array of theoretical and methodological approaches, the collection seeks to further our understanding of how universities can play an effective role in educating graduates suited to the twenty-first century

Psychotherapy and social change:utilizing principles of cognitive-behavioral therapy to help develop new prejudice-reduction interventions

We propose that key concepts from clinical psychotherapy can inform science-based initiatives aimed at building tolerance and community cohesion. Commonalities in social and clinical psychology are identified regarding (1) distorted thinking (intergroup bias and cognitive bias), (2) stress and coping (at intergroup level and intrapersonal level), and (3) anxiety (intergroup anxiety and pathological anxiety). On this basis we introduce a new cognitive-behavioral model of social change. Mental imagery is the conceptual point of synthesis, and anxiety is at the core, through which new treatment-based approaches to reducing prejudice can be developed. More generally, we argue that this integration is illustrative of broader potential for cross-disciplinary integration in the social and clinical sciences, and has the potential to open up new possibilities and opportunities for both disciplines.

Protein modification and phospholipid oxidation

Phospholipid oxidation can generate reactive and electrophilic products that are capable of modifying proteins, especially at cysteine, lysine and histidine residues. Such lipoxidation reactions are known to alter protein structure and function, both with gain of function and loss of activity effects. As well as potential importance in the redox regulation of cell behaviour, lipoxidation products in plasma could also be useful biomarkers for stress conditions. Although studies with antibodies suggested the occurrence of lipoxidation adducts on ApoB-100, these products had not previously been characterized at a molecular level. We have developed new mass spectrometry-based approaches to detect and locate adducts of oxidized phospholipids in plasma proteins, as well as direct oxidation modifications of proteins, which avoid some of the problems typically encountered with database search engines leading to erroneous identifications of oxidative PTMs. This approach uses accurate mass extracted ion chromatograms (XICs) of fragment ions from peptides containing oxPTMs, and allows multiple modifications to be examined regardless of the protein that contains them. For example, a reporter ion at 184.074 Da/e corresponding to phosphocholine indicated the presence of oxidized phosphatidylcholine adducts, while 2 reporter ions at 100.078 and 82.025 Da/e were selective for allysine. ApoB-100-oxidized phospholipid adducts were detected even in healthy human samples, as well as LDL from patients with inflammatory disease. Lipidomic studies showed that more than 350 different species of lipid were present in LDL, and were altered in disease conditions. LDL clearly represents a very complex carrier system and one that offers a rich source of information about systemic conditions, with potential as indicators of oxidative damage in ageing or inflammatory diseases.

Stem cell-derived astrocytes:are they physiologically credible?

Astrocytes are now increasingly acknowledged as having fundamental and sophisticated roles in brain function and dysfunction. Unravelling the complex mechanisms that underlie human brain astrocyte-neuron interactions is therefore an essential step on the way to understanding how the brain operates. Insights into astrocyte function to date, have almost exclusively been derived from studies conducted using murine or rodent models. Whilst these have led to significant discoveries, preliminary work with human astrocytes has revealed a hitherto unknown range of astrocyte types with potentially greater functional complexity and increased neuronal interaction with respect to animal astrocytes. It is becoming apparent, therefore, that many important functions of astrocytes will only be discovered by direct physiological interrogation of human astrocytes. Recent advancements in the field of stem cell biology have provided a source of human based models. These will provide a platform to facilitate our understanding of normal astrocyte functions as well as their role in CNS pathology. A number of recent studies have demonstrated that stem cell derived astrocytes exhibit a range of properties, suggesting that they may be functionally equivalent to their in vivo counterparts. Further validation against in vivo models will ultimately confirm the future utility of these stem-cell based approaches in fulfilling the need for human- based cellular models for basic and clinical research. In this review we discuss the roles of astrocytes in the brain and highlight the extent to which human stem cell derived astrocytes have demonstrated functional activities that are equivalent to that observed in vivo.

Clinical utility of the KAMRA corneal inlay

The treatment of presbyopia has been the focus of much scientific and clinical research over recent years, not least due to an increasingly aging population but also the desire for spectacle independence. Many lens and nonlens-based approaches have been investigated, and with advances in biomaterials and improved surgical methods, removable corneal inlays have been developed. One such development is the KAMRA™ inlay where a small entrance pupil is exploited to create a pinhole-type effect that increases the depth of focus and enables improvement in near visual acuity. Short- and long-term clinical studies have all reported significant improvement in near and intermediate vision compared to preoperative measures following monocular implantation (nondominant eye), with a large proportion of patients achieving Jaeger (J) 2 to J1 (~0.00 logMAR to ~0.10 logMAR) at the final follow-up. Although distance acuity is reduced slightly in the treated eye, binocular visual acuity and function remain very good (mean 0.10 logMAR or better). The safety of the inlay is well established and easily removable, and although some patients have developed corneal changes, these are clinically insignificant and the incidence appears to reduce markedly with advancements in KAMRA design, implantation technique, and femtosecond laser technology. This review aims to summarize the currently published peer-reviewed studies on the safety and efficacy of the KAMRA inlay and discusses the surgical and clinical outcomes with respect to the patient’s visual function.

Handwritten and machine-printed text discrimination using a template matching approach

We propose a novel template matching approach for the discrimination of handwritten and machine-printed text. We first pre-process the scanned document images by performing denoising, circles/lines exclusion and word-block level segmentation. We then align and match characters in a flexible sized gallery with the segmented regions, using parallelised normalised cross-correlation. The experimental results over the Pattern Recognition & Image Analysis Research Lab-Natural History Museum (PRImA-NHM) dataset show remarkably high robustness of the algorithm in classifying cluttered, occluded and noisy samples, in addition to those with significant high missing data. The algorithm, which gives 84.0% classification rate with false positive rate 0.16 over the dataset, does not require training samples and generates compelling results as opposed to the training-based approaches, which have used the same benchmark.