Research Items (23)
This study aimed to investigate the knowledge and attitudes toward communication disorders and speech-language pathology among the general public in Malaysia. A self-developed questionnaire consisting of items related to knowledge and attitudes about communication disorders and speech-language pathology was distributed via Google form and a paper-pencil format. A total of 535 responses were obtained. Descriptive analyses based on the comparison of the demographics to the knowledge and attitude levels were conducted using inferential statistics. More than half of the respondents showed a moderate knowledge level about communication disorders and a high knowledge about speech-language pathology. In terms of attitudes, 67.3% of the respondents perceived a positive attitude towards people with communication disorders, and 86.5% of them were generally positive regarding the profession of speech-language pathology. Health professionals demonstrated a higher knowledge level and more positive attitudes toward communication disorders and speech-language pathology compared to other professions (teachers/services, business, engineers, and jobless/retired). Additionally, respondents with a higher educational level (masters or Ph.D. holders) had greater knowledge about speech-language pathology and more positive attitudes toward communication disorders compared to those who were primary or secondary school graduates. Providing more education and awareness programs about communication disorders and speech-language pathology to the general public should be implemented to improve current practices and expectations in service delivery.
Background To ensure the output quality, current crowdsourcing systems highly rely on redundancy of answers provided by multiple workers with varying expertise, however massive redundancy is very expensive and time-consuming. Task recommendation can help requesters to receive good quality output quicker as well as help workers to find their right tasks faster. To reduce the cost, a number of previous works adopted active learning in crowdsourcing systems for quality assurance. Active learning is a learning approach to achieve certain accuracy with a very low cost. However, previous works do not consider the varying expertise of workers for various task categories in real crowdsourcing scenarios; and they do not consider new workers who are not willing to work on a large amount of tasks before having a list of preferred tasks recommended. In this paper, we propose ActivePMFv2, Probabilistic Matrix Factorization with Active Learning (version 2), on a task recommendation framework called TaskRec to recommend tasks to workers in crowdsourcing systems for quality assurance. By assigning the most uncertain task for new workers to work on, this paper identifies a flaw in our previous ActivePMFv1, Probabilistic Matrix Factorization with Active Learning (version 1). Therefore, ActivePMFv2 can give new workers a list of preferred tasks recommended faster than that of ActivePMFv1. Our factor analysis model considers not only worker task selection preference, but also worker performance history. It actively selects the most uncertain task for the most reliable workers to work on to retrain the classification model. Moreover, we propose a generic online-updating method for learning the model, ActivePMFv2. The larger the profile of a worker (or task) is, the less important is retraining its profile on each new work done. In case of the worker (or task) having large profile, our online-updating algorithm retrains the whole feature vector of the worker (or task) and keeps all other entries in the matrix fixed. Our online-updating algorithm runs batch update to reduce the running time of model update. ResultsComplexity analysis shows that our model is efficient and is scalable to large datasets. Based on experiments on real-world datasets, the result shows that the MAE results and RMSE results of our proposed ActivePMFv2 are improved up to 29 % and 35 % respectively comparing with ActivePMFv1, where ActivePMFv1 outperforms the PMF with other active learning approaches significantly as shown in previous work. Experiment results show that our online-updating algorithm is accurate in approximating to a full retrain of the learning model while the average runtime of model update for each work done is reduced by more than 80 % (decreases from a few minutes to several seconds). Conclusions To the best of our knowledge, we are the first one to use PMF, active learning and dynamic model update to recommend tasks for quality assurance in crowdsourcing systems for real scenarios.
In crowdsourcing systems, task recommendation can help workers to find their right tasks faster as well as help requesters to receive good quality output quicker. A number of previous works adopted active learning for task recommendation in crowdsourcing systems to achieve certain accuracy with a very low cost. However, the model updating methods in previous works are not suitable for real-world applications. In our paper, we propose a generic online-updating method for learning a factor analysis model, ActivePMF on TaskRec (Probabilistic Matrix Factorization with Active Learning on Task Recommendation Framework), for crowdsourcing systems. The larger the profile of a worker (or task) is, the less important is retraining its profile on each new work done. In case of the worker (or task) having large profile, our algorithm only retrains the whole feature vector of the worker (or task) and keeps all other entries in the matrix fixed. Besides, our algorithm runs batch update to further improve the performance. Experiment results show that our online-updating approach is accurate in approximating to a full retrain while the average runtime of model update for each work done is reduced by more than 90 % (from a few minutes to several seconds).
Crowdsourcing is evolving as a distributed problem-solving and business production model in recent years. In crowdsourcing paradigm, tasks are distributed to networked people to complete such that a company’s production cost can be greatly reduced. In crowdsourcing systems, task recommendation can help workers to find their right tasks faster as well as help requesters to receive good quality output quicker. However, previously proposed classification based task recommendation approach, which is the only one in the literature, does not consider the dynamic scenarios of new workers and new tasks in the crowdsourcing system. In this paper, we propose a Task Recommendation (TaskRec) framework based on a unified probabilistic matrix factorization, aiming to recommend tasks to workers in dynamic scenarios. Unlike traditional recommendation systems, workers do not provide their ratings on tasks in crowdsourcing systems, thus we infer user ratings from their interacting behaviors. This conversion helps task recommendation in crowdsourcing systems. Complexity analysis shows that our framework is efficient and is scalable to large datasets. Finally, we conduct experiments on real-world datasets for performance evaluation. Experimental results show that TaskRec outperforms the state-of-the-art approach.
In crowdsourcing systems, task recommendation can help workers to find their right tasks faster as well as help requesters to receive good quality output quicker. However, previously proposed classification approach does not consider the dynamic scenarios of new workers and new tasks in the system. In this paper, we propose a Task Recommendation (TaskRec) framework based on a unified probabilistic matrix factorization, aiming to recommend tasks to workers in dynamic scenarios. Unlike traditional recommendation systems, workers do not provide their ratings on tasks in crowdsourcing systems, and thus we propose to transform worker behaviors into ratings. Complexity analysis shows that our framework is efficient and is scalable to large datasets. Finally, we conduct experiments on real-world datasets for performance evaluation.
In crowdsourcing systems, tasks are distributed to networked people to complete such that a company's production cost can be greatly reduced. Obviously, it is not efficient that the amount of time for a worker spent on selecting a task is comparable with that spent on working on a task, but the monetary reward of a task is just a small amount. The available worker history makes it possible to mine workers' preference on tasks and to provide favorite recommendations. Our exploratory study on the survey results collected from Amazon Mechanical Turk (MTurk) shows that workers' histories can reflect workers' preferences on tasks in crowdsourcing systems. Task recommendation can help workers to find their right tasks faster as well as help requesters to receive good quality output quicker. However, previously proposed classification based task recommendation approach only considers worker performance history, but does not explore worker task searching history. In our paper, we propose a task recommendation framework for task preference modeling and preference-based task recommendation, aiming to recommend tasks to workers who are likely to prefer to work on and provide output that accepted by requesters. We consider both worker performance history and worker task searching history to reflect workers' task preference more accurately. To the best of our knowledge, we are the first to use matrix factorization for task recommendation in crowdsourcing systems.
Crowd sourcing is evolving as a distributed problem-solving and business production model in recent years. In crowd sourcing paradigm, tasks are distributed to networked people to complete such that a company's production cost can be greatly reduced. In 2003, Luis von Ahn and his colleagues pioneered the concept of "human computation", which utilizes human abilities to perform computation tasks that are difficult for computers to process. Later, the term "crowdsourcing" was coined by Jeff Howe in 2006. Since then, a lot of work in crowd sourcing has focused on different aspects of crowd sourcing, such as computational techniques and performance analysis. In this paper, we give a survey on the literature on crowd sourcing which are categorized according to their applications, algorithms, performances and datasets. This paper provides a structured view of the research on crowd sourcing to date.
Crowd sourcing is evolving as a distributed problem-solving and business production model in recent years. In crowd sourcing paradigm, tasks are distributed to networked people to complete such that a company¡¦s production cost can be greatly reduced. A crowd sourcing process involves operations of both requesters and workers. A requester submits a task request, a worker selects and completes a task, and the requester only pays the worker for the successful completion of the task. Obviously, it is not efficient that the amount of time spent on selecting a task is comparable with that spent on working on a task, but the monetary reward of a task is just a small amount. Literature mainly focused on exploring what type of tasks can be deployed to the crowd and analyzing the performance of crowd sourcing platforms. However, no existing work investigates on how to support workers to select tasks on crowd sourcing platforms easily and effectively. In this paper, we propose a novel idea on task matching in crowd sourcing to motivate workers to keep on working on crowd sourcing platforms in long run. The idea utilizes the past task preference and performance of a worker to produce a list of available tasks in the order of best matching with the worker during his task selection stage. It aims to increase the efficiency of task completion. We present some preliminary experimental results in case studies. Finally, we address the possible challenges and discuss the future directions.
Human computation is a technique that makes use of human abilities for computation to solve problems. Social games use the power of the Internet game players to solve human computation problems. In previous works, many social games were proposed and were quite successful, but no formal framework exists for designing social games in general. A formal framework is important because it lists out the design elements of a social game, the characteristics of a human computation problem, and their relationships. With a formal framework, it simplifies the way to design a social game for a specific problem. In this paper, our contributions are: (1) formulate a formal model on social games, (2) analyze the framework and derive some interesting properties based on model's interactions, (3) illustrate how some current social games can be realized with the proposed formal model, and (4) describe how to design a social game for solving a specific problem with the use of the proposed formal model. This paper presents a set of design guidelines derived from the formal model and demonstrates that the model can help to design a social game for solving a specific problem in a formal and structural way.
Human computation is a technique that makes use of human abilities for computation to solve problems. The human computation problems are the problems those computers are not good at solving but are trivial for humans. In this paper, we give a survey of various human computation systems which are categorized into initiatory human computation, distributed human computation and social game-based human computation with volunteers, paid engineers and online players. For the existing large number of social games, some previous works defined various types of social games, but the recent developed social games cannot be categorized based on the previous works. In this paper, we define the categories and the characteristics of social games which are suitable for all existing ones. Besides, we present a survey on the performance aspects of human computation system. This paper gives a better understanding on human computation system.
SIP (Session Initiation Protocol) is a signaling protocol standardized by IETF, aiming to manage the multimedia transmission sessions among different parties. This paper illustrates an adaptive multimedia transmission system for wired and wireless networks based on SIP with protocol selection mechanism for a certain level of QoS guarantee. In our system, SIP is not only used for call setup signaling but also for carrying the information in the protocol selection. Using Agent Server, our system can select the most suitable protocol for adapting different situations intelligently during the connections and data buffering service is also provided for various media data flows between the end users with acceptable QoS level without any interruption and disconnection regardless of types of devices, platforms and protocols used.
This paper proposes a new distributed QoS routing protocol, called Efficient Distributed QoS Routing (EDQR), for MPLS networks. The path searching algorithm of EDQR considers an additional parameter to select the optimal path from limited choices, and its implementation requires both source and destination of a connection request to make the routing decision. Our simulation results show that, the operation cost of EDQR is much lower than other popular QoS routing protocols but with very similar network performance. Especially, the network performance of our protocol is better when the network loading is high. Moreover, we develop a very simple mathematical model to investigate the upper bound of the size of limited choices for engineering design.
In order to support real-time video, audio and data communication among heterogeneous third generation (3G) handsets, 3G phones/terminals are required to support 3G-324M, the multimedia transmission protocol stack for 3G communication. This paper discusses some efficient approaches and experiences in the implementation of 3G-324M protocol stack. Specifically, we discuss: (1) event-driven approach for the overall information exchange; (2) single-step direct message transformation for the optimization of tree-structured message processing and (3) serialization of nested multiplex table entries in multiplexing/demultiplexing processing. Our implementation has been tested in a realistic heterogeneous 3G communication environment for transmission of real-time video, audio and data and its performance is satisfactory.
3G-324M is a multimedia transmission protocol designed for 3G communication environment. Meanwhile H.245 standard is a control protocol in 3G-324M and gives specific descriptions about terminal information messages in H.245 control channel as well as the procedures using them. The message syntax is defined using an external data representation standard called Abstract Syntax Notation One (ASN.1). For transmission, ASN.1 formatted data is transformed into bit-stream based on an ASN.1 encoding standard called Packed Encoding Rules (PER). In order to meet the requirement of high speed data transfer in 3G communication, it is important to design the procedure of message processing as simple as possible. In this paper, we propose Single-step Direct Message Transformation (SDMT)for the optimization of tree-structured message processing in H.245 module. By testing in realistic environments in some China industries, performance evaluation shows that code redundancies in terms of file size and code size are reduced significantly.
Context-aware computing is a computing paradigm in which applications can take advantage of contextual information. Quality of network connection is a very important factor for mobile web services. However, the conditions of mobile networks may change frequently and dynamically. Thus, providing support for context-aware applications is especially important in mobile web services. Recently, a number of architectures supporting context-aware applications have been developed, but little attention is paid to the special requirements of mobile devices which particularly have many constraints. This paper discusses a client-proxy-server architecture that supports context-awareness by considering types of device, network and application characteristics. The contribution of this paper mainly lies in the division of labor between proxy and server. Application specific proxy is used to tailor the original resource based on the mobile user’s context information. To prove the feasibility, a context-aware image management system is designed and realized.
Multimedia transmission over wireless networks is a hot research field, but little work focuses on discussing efficient transmission between wireless terminals. The limited bandwidth of wireless networks and the limited capabilities of wireless terminals are two major problems for the efficient transmission of multimedia flows. We propose an end-to-end wireless multimedia transmission (EEWMT) system that, as well as supporting end-to-end wireless transmission for multimedia flows, provides services to resolve the above two problems. The EEWMT system is mainly made up of three parts: anyDevice, which is a protocol stack working on terminals; a center server, which provides indexing and database services and is responsible for session management; an agent server, which provides data buffering and multimedia transcoding services. Implementation of our system shows that it enhances both the communication efficiency and the content quality.
We develop a very simple mathematical model to investigate the upper bound of the size of limited choices in our proposed QoS path selection algorithm called the largest widest shortest path among limited choices (LWSP-LC) for engineering design. The LWSP-LC can achieve load balancing effectively and reduce path searching complexity significantly. However, its performance is highly related to many factors, including the network environment and traffic condition. Our simulation results show that the size of limited choices is highly dependent on both the network size and network connectivity. By considering a fully connected network, we derive a simple mathematical model on the upper bound of the size of limited choices based on the network size only.
This paper presents a novel Generic Communication Model (GCM) for integrated mobile and Internet web services. GCM is an adaptive protocol selection mechanism to ease the balance of transmission performance and communication interoperability in wired networks and wireless mobile net- works. With GCM, end users can select a suitable protocol among different pro- tocols for adapting different situations intelligently without any interruption and disconnection. In this way, it can provide seamless mobile communication by integrating both 3G telecom networks and IEEE 802.11 wireless networks through GCM which allow various client devices to communicate at their best acceptable QoS level through consistent and simple APIs. We present a pre- liminary prototype of GCM to illustrate its functionalities of communicating with heterogeneous mobile devices by using various communication protocols.
- Oct 2003
- Communications, 2003. APCC 2003. The 9th Asia-Pacific Conference on
In this paper, we propose a new efficient QoS path searching algorithm called the largest widest shortest path with limited choices (LWSP-LC) for load balancing in the Internet. This algorithm is modified from the widest shortest path (WSP) with two important modifications: our algorithm considers an addition parameter in the path selection criteria and searches the optimal path from very limited choices. By comparing with the WSP, our simulation results show that the LWSP-LC has a lower computational complexity, which is up to 100 times less than the WSP, without any performance degradation.
In this paper, we propose a new algorithm, called Dynamic DNS for Load Balancing (DDLB), embedded in DNS (Domain Name Service) servers. By applying our algorithm into a simulation model with real data and comparing its performance with the original round-robin DNS setting, we found that around 40% of the incoming requests had shorter service time and almost no incoming requests got longer service time. We have also applied DDLB into a real system and we found that the improvement was up to 58% in the peak period. Our proposed algorithm not only works very well in a real system, but also has simple implementation such that the existing configuration can be easily modified to apply.