Open Access

Semantic classification of words using distributional features is usually based on the semantic similarity of words. We show on two different datasets that a trained classifier using the distributional features directly gives better results. We use Support Vector Machines (SVM) and Multirelational Matrix Factorization (MRMF) to train classifiers. Both give similar results. However, MRMF, that was not used for semantic classification with distributional features before, can easily be extended with more matrices containing more information from different sources on the same problem. We demonstrate the effectiveness of the novel approach by including information from WordNet. Thus we show, that MRMF provides an interesting approach for building semantic classifiers that (1) gives better results than unsupervised approaches based on vector similarity, (2) gives similar results as other supervised methods and (3) can naturally be extended with other sources of information in order to improve the results.

In distributional semantics words are represented by aggregated context features. The similarity of words can be computed by comparing their feature vectors. Thus, we can predict whether two words are synonymous or similar with respect to some other semantic relation. We will show on six different datasets of pairs of similar and non-similar words that a supervised learning algorithm on feature vectors representing pairs of words outperforms cosine similarity between vectors representing single words. We compared different methods to construct a feature vector representing a pair of words. We show that simple methods like pairwise addition or multiplication give better results than a recently proposed method that combines different types of features. The semantic relation we consider is relatedness of terms in thesauri for intellectual document classification. Thus our findings can directly be applied for the maintenance and extension of such thesauri. To the best of our knowledge this relation was not considered before in the field of distributional semantics.

Background: After kidney transplantation, immunosuppressive therapy causes impaired cellular immune defense leading to an increased risk of viral complications. Trough level monitoring of immunosuppressants is insufficient to estimate the individual intensity of immunosuppression. We have already shown that virus-specific T cells (Tvis) correlate with control of virus replication as well as with the intensity of immunosuppression. The multicentre IVIST01-trial should prove that additional steering of immunosuppressive and antiviral therapy by Tvis levels leads to better graft function by avoidance of over-immunosuppression (for example, viral infections) and drug toxicity (for example, nephrotoxicity). Methods/design: The IVIST-trial starts 4 weeks after transplantation. Sixty-four pediatric kidney recipients are randomized either to a non-intervention group that is only treated conservatively or to an intervention group with additional monitoring by Tvis. The randomization is stratified by centre and cytomegalovirus (CMV) prophylaxis. In both groups the immunosuppressive medication (cyclosporine A and everolimus) is adopted in the same target range of trough levels. In the non-intervention group the immunosuppressive therapy (cyclosporine A and everolimus) is only steered by classical trough level monitoring and the antiviral therapy of a CMV infection is performed according to a standard protocol. In contrast, in the intervention group the dose of immunosuppressants is individually adopted according to Tvis levels as a direct measure of the intensity of immunosuppression in addition to classical trough level monitoring. In case of CMV infection or reactivation the antiviral management is based on the individual CMV-specific immune defense assessed by the CMV-Tvis level. Primary endpoint of the study is the glomerular filtration rate 2 years after transplantation; secondary endpoints are the number and severity of viral infections and the incidence of side effects of immunosuppressive and antiviral drugs. Discussion: This IVIST01-trial will answer the question whether the new concept of steering immunosuppressive and antiviral therapy by Tvis levels leads to better future graft function. In terms of an effect-related drug monitoring, the study design aims to realize a personalization of immunosuppressive and antiviral management after transplantation. Based on the IVIST01-trial, immunomonitoring by Tvis might be incorporated into routine care after kidney transplantation.

The network security framework VisITMeta allows the visual evaluation and management of security event detection policies. By means of a "what-if" simulation the sensitivity of policies to specific events can be tested and adjusted. This paper presents the results of a user study for testing the usability of the approach by measuring the correct completion of given tasks as well as the user satisfaction by means of the system usability scale.

Objective The study’s objective was to assess factors contributing to the use of smart devices by general practitioners (GPs) and patients in the health domain, while specifically addressing the situation in Germany, and to determine whether, and if so, how both groups differ in their perceptions of these technologies. Methods GPs and patients of resident practices in the Hannover region, Germany, were surveyed between April and June 2014. A total of 412 GPs in this region were invited by email to participate via an electronic survey, with 50 GPs actually doing so (response rate 12.1%). For surveying the patients, eight regional resident practices were visited by study personnel (once each). Every second patient arriving there (inclusion criteria: of age, fluent in German) was asked to take part (paper-based questionnaire). One hundred and seventy patients participated; 15 patients who did not give consent were excluded. Results The majority of the participating patients (68.2%, 116/170) and GPs (76%, 38/50) owned mobile devices. Of the patients, 49.9% (57/116) already made health-related use of mobile devices; 95% (36/38) of the participating GPs used them in a professional context. For patients, age (P < 0.001) and education (P < 0.001) were significant factors, but not gender (P > 0.99). For doctors, neither age (P = 0.73), professional experience (P > 0.99) nor gender (P = 0.19) influenced usage rates. For patients, the primary use case was obtaining health (service)-related information. For GPs, interprofessional communication and retrieving information were in the foreground. There was little app-related interaction between both groups. Conclusions GPs and patients use smart mobile devices to serve their specific interests. However, the full potentials of mobile technologies for health purposes are not yet being taken advantage of. Doctors as well as other care providers and the patients should work together on exploring and realising the potential benefits of the technology.