Activity Modeling Using Semantic-Based Reasoning to Provide Meaningful Context in Human Activity Recognizing

Amnal, AnisRahmawati; Dewabharata, Anindhita; Chou, Shou-Yan; Erawan, Mahendrawathi

doi:10.1007/978-981-4451-98-7_129

Cited by 1 publication

(3 citation statements)

References 7 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For instance, in Fig.4, the experiment results show that the proposed context inconsistency elimination algorithm improves the context aware rate by up to 3.9% compared to the trust-worthiness based method, 2.2% compared to the voting method and 2.4% compared to the efficient context-aware selection based on user feedback method when the error rate 1 E is equal to 0.2.…”

Section: A Ds Fdmentioning

confidence: 99%

“…It is assumed that the perception error rates of sensors 2 3 4 5 , , , S S S S are changeless within each experiment, but the perception error rates of sensor 1 S vary from small to large. Fig.…”

Section: The Performance Evaluationmentioning

confidence: 99%

“…Context aware technology is one of the key technologies of ubiquitous computing [1], [2]. Various sensors can be applied to deriving context.…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Context Inconsistency Elimination Based on User Feedback and Modified Evidence Theory

Wang¹,

Xu²,

Sun³

et al. 2013

IJMLC

View full text Add to dashboard Cite

Abstract-In dynamic and open environment, context aware system obtains context information from the dynamic, distributed and heterogeneous sources, but the context information usually has inconsistencies which would lead to inappropriate services. We proposed a new context inconsistency elimination algorithm based on user feedback and modified evidence theory in this paper. Through user feedback, each sensor's perception precision can be acquired, and with the modified evidence theory, we can make full use of all context information and eliminate inconsistent context by adjusting the influence of every context on whole judgment based on sensor perception precision. In order to evaluate the performance of the proposed context inconsistency elimination algorithm, context aware rate is defined. The experiment results show that the proposed context inconsistency elimination algorithm can obtain the best context aware rate in most cases when the error rates of sensors are varied. Index Terms-Adaptive service, context aware, context inconsistency elimination, user feedback I. INTRODUCTIONContext aware technology is one of the key technologies of ubiquitous computing [1], [2]. Various sensors can be applied to deriving context. Through the consideration of contexts, the context aware systems are able to adapt their services to the continuously evolving situations without diverting human attention so as to improve the quality of human-computer interaction experience.However, due to the precision differences of sensors, equipment isomerism and network delay, the context aware information usually has inconsistencies [3]. Thus, we should eliminate those inconsistencies so as to obtain the correct context information [4]. Otherwise the services we provide would be inappropriate.In [5], the authors present the trust-worthiness based algorithm which selects the context information with the largest sensor perception precision among all context information. In [6] the voting algorithm is applied to select the context information by plurality. In addition, efficient context aware selection based on user feedback algorithm which takes Manuscript received September 20, 2013; revised November 18, 2013 trust-worthiness based algorithm [5] and voting based algorithm [6] into consideration is proposed in [7]. In this algorithm, the authors set accepted threshold and rejection threshold. If there is a sensor perception precision which is larger than accepted threshold, the service provider will select the context information of this sensor as final output; or if the sensor perception precisions of all sensors are under the rejection threshold, the trust-worthiness based algorithm will be used; otherwise, voting algorithm would be applied. Nevertheless, above mentioned algorithms cannot make full use of all context information, so they could not obtain outstanding performance.In [8]-[10], the parameter of Quality of Context (QoC) in middleware is used to manage context information. In [11], QoC is measured on context source layer, conte...

show abstract

Section: A Ds Fdmentioning

confidence: 99%

Section: The Performance Evaluationmentioning

confidence: 99%