Speech-based solution to Parkinson’s disease management

Parkinson’s disease (PD) is a neurological disorder marked by decreased dopamine levels in the brain. Persons suffering from PD, exhibits vocal symptoms such as dysphonia and dysarthria. Speech impairments in PD are grouped together and called as hypokinetic dysarthria. Traditional PD management is based on a patient’s clinical history and through physical examination as there are currently no known biomarkers for its diagnosis. Automatic analysis techniques aid clinicians in diagnosis and monitoring patients using speech and provide frequent, cost effective and objective assessment. This paper presents pilot experiment to detect presence of dysarthria in speech and detect level of severity based on deep learning approach. Automated feature extraction and classification using convolutional neural network shows 77.48% accuracy on test samples of TORGO database with five fold validation. Using transfer learning, system performance is further analyzed for gender specific performance as well as in detection of severity of disease.

     

DYRE: a DYnamic REconfigurable solution to increase GPGPU’s reliability

The Journal of Supercomputing - General-purpose graphics processing units (GPGPUs) are extensively used in high-performance computing. However, it is well known that these devices’...     

The influence of multimedia training on users’ attitudes: lessons learned

The workforce within the US is increasingly culturally diverse and mobile. Rapid technological changes, coupled with cultural diversity and employee mobility, have created the milieu for critical issues in organizational training methods. Effective learning systems enable employees to adapt more easily to change, thereby increasing their effectiveness. Multimedia systems can be the solution to enhancing performance effectively and efficiently by accelerated learning. This paper examines such a system developed by Holiday Inn Worldwide for the implementation of a new information system. Individuals’ performance and attitudes were tracked for two primary groups: those receiving training on a new information system using traditional training methods and those using a multimedia CDROM to acquire mastery of the same system. 826 employees participated, 467 of them were trained by the multimedia system. Attitude measures included attitudes toward self: job, training, and technology. Individual performance was measured by standardized tests on employees’ knowledge of the new information system. Our findings demonstrated that users’ attitudes were consistently poorer when trained at sites which employed multimedia training when compared with users trained in using a more traditional approach.     

Usability of a cloud-based collaborative learning framework to improve learners’ experience

Computer-Supported Collaborative Learning (CSCL) is concerned with how Information and Communication Technology (ICT) might facilitate learning in groups which can be co-located or distributed over a network of computers such as Internet. CSCL supports effective learning by means of communication of ideas and information among learners, collaborative access of essential documents, and feedback from instructors and peers on learning activities. As the cloud technologies are increasingly becoming popular and collaborative learning is evolving, new directions for development of collaborative learning tools deployed on cloud are proposed. Development of such learning tools requires access to substantial data stored in the cloud. Ensuring efficient access to such data is hindered by the high latencies of wide-area networks underlying the cloud infrastructures. To improve learners’ experience by accelerating data access, important files can be replicated so a group of learners can access data from nearby locations. Since a cloud environment is highly dynamic, resource availability, network latency, and learner requests may change. In this paper, we present the advantages of collaborative learning and focus on the importance of data replication in the design of such a dynamic cloud-based system that a collaborative learning portal uses. To this end, we introduce a highly distributed replication technique that determines optimal data locations to improve access performance by minimizing replication overhead (access and update). The problem is formulated using dynamic programming. Experimental results demonstrate the usefulness of the proposed collaborative learning system used by institutions in geographically distributed locations.     

Exploiting users’ social relations to forward data in opportunistic networks: The HiBOp solution

Opportunistic networks, in which nodes opportunistically exploit any pair-wise contact to identify next hops towards the destination, are one of the most interesting technologies to support the pervasive networking vision. Opportunistic networks allow content sharing between mobile users without requiring any pre-existing Internet infrastructure, and tolerate partitions, long disconnections, and topology instability in general. In this paper we propose a context-aware framework for routing and forwarding in opportunistic networks. The framework is general, and able to host various flavors of context-aware routing. In this work we also present a particular protocol, HiBOp, which, by exploiting the framework, learns and represents through context information, the users’ behavior and their social relations, and uses this knowledge to drive the forwarding process. The comparison of HiBOp with reference to alternative solutions shows that a context-aware approach based on users’ social relations turns out to be a very efficient solution for forwarding in opportunistic networks. We show performance improvements over the reference solutions both in terms of resource utilization and in terms of user perceived QoS.     

The night float paradigm to decrease sleep deprivation: good solution or a new problem?

In the late 1980s physician residency training programs developed the night float rotation, characterized by a sequence of 5 - 15 days of night work without any daytime duties, thereby involving an abrupt reversal of the wake-sleep schedule. We examined the effect of the night float rotation on sleep, mood and performance of pediatric residents. Residents completed sleep diaries daily, and tests of mood (Profile of Mood States) and attention (Conner's Continuous Performance Test) three times a week during the two-week night float rotation, and during equivalent blocks of time of their daytime rotations. Results show that, despite having ample opportunity to sleep during the day, while on night float rotation residents slept less than during the nights of their normal daytime rotations, 6.3 h ± 2.5 h and 7.2 h ± 1.7 h, respectively, p < 0.0001. Also, during night float compared to daytime rotations residents had increased fatigue-inertia scores, 8.7 ± 4.1 and 4.8 ± 2.4, respectively, p < 0.0001, and decreased vigor-activity scores 10.7 ± 5.4 and 14.8 ± 5.3, respectively, p = 0.02. The scores for attention were not significantly different between night float and daytime rotations. The correlation coefficients of fatigue with measures of attention were not statistically significant for daytime rotations. However, for night float fatigue correlated with omission errors, r = 0.51, p = 0.001 and with attentiveness r = ?0.36, p = 0.03. Training programs that adopt the night float rotation must be aware of potential deleterious effects of the night float rotation as they may lead to serious consequences on residents' performance and patients' safety.     

The night float paradigm to decrease sleep deprivation: good solution or a new problem?

In the late 1980s physician residency training programs developed the night float rotation, characterized by a sequence of 5 - 15 days of night work without any daytime duties, thereby involving an abrupt reversal of the wake - sleep schedule. We examined the effect of the night float rotation on sleep, mood and performance of pediatric residents. Residents completed sleep diaries daily, and tests of mood (Profile of Mood States) and attention (Conner's Continuous Performance Test) three times a week during the two-week night float rotation, and during equivalent blocks of time of their daytime rotations. Results show that, despite having ample opportunity to sleep during the day, while on night float rotation residents slept less than during the nights of their normal daytime rotations, 6.3 h +/- 2.5 h and 7.2 h +/- 1.7 h, respectively, p < 0.0001. Also, during night float compared to daytime rotations residents had increased fatigue-inertia scores, 8.7 +/- 4.1 and 4.8 +/- 2.4, respectively, p < 0.0001, and decreased vigor-activity scores 10.7 +/- 5.4 and 14.8 +/- 5.3, respectively, p = 0.02. The scores for attention were not significantly different between night float and daytime rotations. The correlation coefficients of fatigue with measures of attention were not statistically significant for daytime rotations. However, for night float fatigue correlated with omission errors, r = 0.51, p = 0.001 and with attentiveness r = - 0.36, p = 0.03. Training programs that adopt the night float rotation must be aware of potential deleterious effects of the night float rotation as they may lead to serious consequences on residents' performance and patients' safety.     

An Indian solution to ‘incompleteness’

Kurt Gödel’s Incompleteness theorem is well known in Mathematics/Logic/Philosophy circles. Gödel was able to find a way for any given P (UTM), (read as, “P of UTM” for “Program of Universal Truth Machine”), actually to write down a complicated polynomial that has a solution iff (=if and only if), G is true, where G stands for a Gödel-sentence. So, if G’s truth is a necessary condition for the truth of a given polynomial, then P (UTM) has to answer first that G is true in order to secure the truth of the said polynomial. But, interestingly, P (UTM) could never answer that G was true. This necessarily implies that there is at least one truth a P (UTM), however large it may be, cannot speak out. Daya Krishna and Karl Potter’s controversy regarding the construal of India’s Philosophies dates back to the time of Potter’s publication of “Presuppositions of India’s Philosophies” (1963, Englewood Cliffs Prentice-Hall Inc.) In attacking many of India’s philosophies, Daya Krishna appears to have unwittingly touched a crucial point: how can there be the knowledge of a ‘non-cognitive’ mok?a? [‘mok?a’ is the final state of existence of an individual away from Social Contract]—See this author’s Indian Social Contract and its Dissolution (2008) mok?a does not permit the knowledge of one’s own self in the ordinary way with threefold distinction, i.e., subject–knowledge-object or knower–knowledge–known. But what is important is to demonstrate whether such ‘knowledge’ of non-cognitive mok?a state can be logically shown, in a language, to be possible to attain, and that there is no contradiction involved in such demonstration, because, no one can possibly express the ‘experience-itself’ in language. Hence, if such ‘knowledge’ can be shown to be logically not impossible in language, then, not only Daya Krishna’s arguments against ‘non-cognitive mok?a’ get refuted but also it would show the possibility of achieving ‘completeness’ in its truest sense, as opposed to Gödel’s ‘Incompleteness’. In such circumstances, man would himself become a Universal Truth Machine. This is because the final state of mok?a is construed as the state of complete knowledge in Advaita. This possibility of ‘completeness’ is set in this paper in the backdrop of ?rī ?a?karācārya’s Advaitic (Non-dualistic) claim involved in the mahāvākyas (extra-ordinary propositions). (Mahāvākyas that ?a?kara refers to are basically taken from different Upani?ads. For example, “Aham Brahmāsmi” is from B?hadāra?yaka Upanisad, and “Tattvamasi” is from Chāndogya Upani?ad. ?rī ?a?karācārya has written extensively. His main works include his Commentary on Brahma-Sūtras, on major Upani?ads, and on ?rīmadBhagavadGītā, called Bhā?yas of them, respectively. Almost all these works are available in English translation published by Advaita Ashrama, 5 Dehi Entally Road, Calcutta, 700014.) On the other hand, the ‘Incompleteness’ of Gödel is due to the intervening G-sentence, which has an adverse self-referential element. Gödel’s incompleteness theorem in its mathematical form with an elaborate introduction by R.W. Braithwaite can be found in Meltzer (Kurt Gödel: on formally undecidable propositions of principia mathematica and related systems. Oliver &; Boyd, Edinburgh, 1962). The present author believes first that semantic content cannot be substituted by any amount of arithmoquining, (Arithmoquining or arithmatization means, as Braithwaite says,—“Gödel’s novel metamathematical method is that of attaching numbers to the signs, to the series of signs (formulae) and to the series of series of signs (“proof-schemata”) which occur in his formal system…Gödel invented what might be called co-ordinate metamathematics…”) Meltzer (1962 p. 7). In Antone (2006) it is said “The problem is that he (Gödel) tries to replace an abstract version of the number (which can exist) with the concept of a real number version of that abstract notion. We can state the abstraction of what the number needs to be, [the arithmoquining of a number cannot be a proof-pair and an arithmoquine] but that is a concept that cannot be turned into a specific number, because by definition no such number can exist.”.), especially so where first-hand personal experience is called for. Therefore, what ultimately rules is the semanticity as in a first-hand experience. Similar points are voiced, albeit implicitly, in Antone (Who understands Gödel’s incompleteness theorem, 2006). (“…it is so important to understand that Gödel’s theorem only is true with respect to formal systems—which is the exact opposite of the analogous UTM (Antone (2006) webpage 2. And galatomic says in the same discussion chain that “saying” that it ((is)) only true for formal systems is more significant… We only know the world through “formal” categories of understanding… If the world as it is in itself has no incompleteness problem, which I am sure is true, it does not mean much, because that is not the world of time and space that we experience. So it is more significant that formal systems are incomplete than the inexperiencable ‘World in Itself’ has no such problem.—galatomic”) Antone (2006) webpage 2. Nevertheless galatomic certainly, but unwittingly succeeds in highlighting the possibility of experiencing the ‘completeness’ Second, even if any formal system including the system of Advaita of ?a?kara is to be subsumed or interpreted under Gödel’s theorem, or Tarski’s semantic unprovability theses, the ultimate appeal would lie to the point of human involvement in realizing completeness since any formal system is ‘Incomplete’ always by its very nature as ‘objectual’, and fails to comprehend the ‘subject’ within its fold.     

Multimedia teleorchestra with independent sources: Part 1 — temporal modeling of collaborative multimedia scenarios

Multimedia synchronization involving independent sources is a challenging issue imposed by the distributed multimedia applications. In our work, this issue is studied by investigating the teleorchestra application (remote multimedia presentation). In teleorchestration, among the data objects to be presented, relative and uncertain temporal requirements may be involved. Fuzzy presentation scenarios are thus generated. In this paper, we describe a temporal model that can handle these fuzzy scenarios that contain imprecise synchronization constraints, such as unknown object presentation durations and relative event occurring times. The model supports a distributed synchronization algorithm that can schedule the independent sources for the multimedia teleorchestration.     

The effect of ‘ unwanted’ signals on performance in a vigilance task: a reply to Jerison

Jerison's remarks (Ergonomics, 1966, 9, 413) are discussed in relation to the results of a further experiment on ' unwanted ' signals in which (a) the original interaction between the effects of signal probability and signal disc location was not observed, and (b) it was demonstrated that a search requirement is not a necessary condition for eliciting the main effects previously found.     

The drivers of consumers’ intention to redeem a push mobile coupon

A modern-day marketing tool that has caught the industry's attention is the use of mobile coupons (m-coupons). Despite the attractiveness of this new marketing tool, a high level of consumer resistance is reported. To therefore ensure the successful implementation of an m-coupon strategy, customer buy-in is a prerequisite. The research on which this article is based, aimed to explore the factors that contribute to consumers’ intention to ultimately redeem m-coupons. A hypothesised model proposes economic benefit, convenience benefit, positive consumer attitude, perceived control and social benefit as constructs that either directly or indirectly influence consumers’ intention to redeem an m-coupon. A survey of 204 respondents revealed that a consumer's positive attitude is the main driving force behind m-coupon redemption intentions. Results reflected that businesses should emphasise convenience and economic benefits in order to assist in the development of a positive attitude amongst potential m-coupon users. Consumers value their privacy and are resistant towards push-based m-coupons. If businesses incorporate these aspects into their m-couponing strategy, it will allow them to reach their audience in an entirely new manner and in doing so, potentially trigger purchase behaviour or increase store foot count.     

Transferring auditors’ internal control evaluation knowledge to management

Due to both regulatory and competitive forces, attention to an organization's internal controls has increased significantly in the 1990s. Although management is ultimately responsible for ensuring internal controls are adequate, managers often lack knowledge of internal control concepts. This study reports on an experiment testing an expert system developed to facilitate the transfer of internal control knowledge to management. Experimental results indicate that expert systems are viable aids for transferring internal control knowledge to managers whose work experience is outside of accounting and control systems.     

Student data mining solution–knowledge management system related to higher education institutions

Higher education institutions (HEIs) are often curious whether students will be successful or not during their study. Before or during their courses the academic institutions try to estimate the percentage of successful students. But is it possible to predict the success rate of students enrolled in their courses? Are there any specific student characteristics, which can be associated with the student success rate? Is there any relevant student data available to HEIs on the basis of which they could predict the student success rate? The answers to the above research questions can generally be obtained using data mining tools. Unfortunately, data mining algorithms work best with large data sets, while student data, available to HEIs, related to courses are limited and falls into the category of small data sets. Thus, the study focuses on data mining for small student data sets and aims to answer the above research questions by comparing two different data mining tools. The conclusions of this study are very promising and will encourage HEIs to incorporate data mining tools as an important part of their higher education knowledge management systems.     

‘DreamTeam’: A platform for synchronous collaborative applications

This paper presents a platform for developing, testing and executing synchronous collaborative applications in a distributed, heterogeneous environment. Even though several environments exist nowadays, specific problems are not treated satisfactorily. Especially in real network environments, problems like unstable network connections and low bandwidths have to be considered.The DreamTeam platform addresses the special needs of environments with non-optimal characteristics which can, be found in distance learning scenarios. DreamTeam comprises a development environment, a simulation environment and a runtime environment; it is based upon the concept of a fully decentralised architecture and encourages rapid prototyping.DreamTeam supports developers of shared applications through a component concept. Using components helps to divide a software project into well-defined parts. Well-documented interfaces help to reduce integration efforts and improve software quality. A selection of sample applications with DreamTeam validates our design concept.This paper is an extended version of Roth, J. and Unger, C. (1998). DreamTeam: A Platform for Synchronous Collaborative Applications. In Herrmann, Th. and Just-Hahn, K. (eds)Groupware und organisatorische Innovation (D-CSCW '98). B.G. Teubner, Stuttgart, 153–165.