Search

In this paper, a systematic review of relevant published studies on computer-based speech therapy systems or virtual speech therapists (VSTs) for people with speech disorders is presented. We structured this work based on the PRISMA framework. The advancements in speech technology and the increased number of successful real-world projects in this area point to a thriving market for VSTs in the near future; however, there is no standard roadmap to pinpoint how these systems should be designed, implemented, customized, and evaluated with respect to the various speech disorders. The focus of this systematic review is on articulation and phonological impairments. This systematic review addresses three research questions: what types of articulation and phonological disorders do VSTs address, how effective are virtual speech therapists, and what technological elements have been utilized in VST projects. The reviewed papers were sourced from comprehensive digital libraries, and were published in English between 2004 and 2014. All the selected studies involve computer-based intervention in the form of a VST regarding articulation or phonological impairments, followed by qualitative and/or quantitative assessments. To generate this review, we encountered several challenges. Studies were heterogeneous in terms of disorders, type and frequency of therapy, sample size, level of functionality, etc. Thus, overall conclusions were difficult to draw. Commonly, publications with rigorous study designs did not describe the technical elements used in their VST, and publications that did describe technical elements had poor study designs. Despite this heterogeneity, the selected studies reported the effectiveness of computers as a more engaging type of intervention with more tools to enrich the intervention programs, particularly when it comes to children; however, it was emphasized that virtual therapists should not drive the intervention but must be used as a medium to deliver the intervention planned by speech-language pathologists. Based on the reviewed papers, VSTs are significantly effective in training people with a variety of speech disorders; however, it cannot be claimed that a consensus exists in the superiority of VSTs over speech-language pathologists regarding rehabilitation outcomes. Our review shows that hearing-impaired cases were the most frequently addressed disorder in the reviewed studies. Automatic speech recognition, speech corpus, and speech synthesizers were the most popular technologies used in the VSTs. [ABSTRACT FROM AUTHOR]

Cadmium-enrichment of farmland soil greatly threatens the sustainable use of soil resources and the safe cultivation of grain. This review paper briefly introduces the status of farmland soil as well as grain, which are both often polluted by cadmium (Cd) in China, and illustrates the major sources of Cd contaminants in farmland soil. In order to meet soil environmental quality standards and farmland environmental quality evaluation standards for edible agricultural products, Cd-enriched farmland soil is frequently remediated with the following prevailing techniques: dig and fill, electro-kinetic remediation, chemical elution, stabilisation and solidification, phytoremediation, field management and combined remediation. Most remediation techniques are still at the stage of small-scale trial experiments in China and few techniques are assessed in field trials. After comparing the technical and economical applicability among different Cd-enriched farmland soil remediation techniques, a novel ecological and hydraulic remediation technique has been proposed, which integrated the advantages of chemical elution, solidification and stabilisation, phytoremediation and field management. The ecological and hydraulic remediation concept is based on existing irrigation and drainage facilities, ecological ditches (ponds) and agronomic measures, which mainly detoxify the Cd-enriched soil during the interim period of crop cultivation, and guarantee the grain safety during its growth period. This technique may shift the challenge from soil to water treatment, and thus greatly enhances the remediation efficiency and shortens the remediation duration. Moreover, the proposed ecological and hydraulic remediation method matches well with the practical choice of cultivation while remediation for Cd-enriched soil in China, which has negligible impacts on the normal crop cultivation process, and thus shows great potential for large area applications. [ABSTRACT FROM AUTHOR]

Systematic reviews require expert reviewers to manually screen thousands of citations in order to identify all relevant articles to the review. Active learning text classification is a supervised machine learning approach that has been shown to significantly reduce the manual annotation workload by semi-automating the citation screening process of systematic reviews. In this paper, we present a new topic detection method that induces an informative representation of studies, to improve the performance of the underlying active learner. Our proposed topic detection method uses a neural network-based vector space model to capture semantic similarities between documents. We firstly represent documents within the vector space, and cluster the documents into a predefined number of clusters. The centroids of the clusters are treated as latent topics. We then represent each document as a mixture of latent topics. For evaluation purposes, we employ the active learning strategy using both our novel topic detection method and a baseline topic model (i.e., Latent Dirichlet Allocation). Results obtained demonstrate that our method is able to achieve a high sensitivity of eligible studies and a significantly reduced manual annotation cost when compared to the baseline method. This observation is consistent across two clinical and three public health reviews. The tool introduced in this work is available from https://nactem.ac.uk/pvtopic/. [ABSTRACT FROM AUTHOR]