Mahi Lohi

A Comparative Study of Selected Classifiers with Classification Accuracy in User Profiling

In recent years the use of personalized service provisioning applications has been very popular. However, effective personalization cannot be achieved without accurate user profiles. In literature a number of classification algorithms have been used to classify user related information to create accurate user profiles. Nevertheless, there is lack of comparison of these algorithms with classification accuracy of the user profile information. In our previous work [1], we compared four different classification algorithms which are; Naïve Bayesian (NB), Instance-Based Learner (IB1), Bayesian networks (BN) and Lazy Learning of Bayesian Rules (LBR) classifiers. According to our results NB and IB1 classifiers outperformed the BN and LBR classifiers with respect to classification accuracy. In this study we compare the performance of NB, IB1, Classification and Regression Tree (SimpleCART), Naïve Bayesian Tree (NBTree), Iterative Dichotomister Tree (Id3), J48 -a version of C4.5- and Sequential Minimal Optimization (SMO) algorithms with large user profile data. This study is aimed to find the best classification algorithm for user profiling process.Our simulation results show that, in general, the NBTree has the highest classification accuracy performance with the lowest error rate. On the other hand, we also found that the NBTree has one of the highest time requirements to build the classification model. Therefore, NBTree classification algorithm should be favoured over SMO, NB, IB1, J48, SimpleCART and Id3 classifiers in the personalization applications especially when the classification accuracy performance is important.

Classification accuracy performance of Naïve Bayesian (NB), Bayesian Networks (BN), Lazy Learning of Bayesian Rules (LBR) and Instance-Based Learner (IB1) - comparative study

In recent years the used of personalization in service provisioning applications has been very popular. However, effective personalization cannot be achieved without accurate user profiles. A number of classification algorithms have been used to classify user related information to create accurate user profiles. In this study four different classification algorithms which are; naive Bayesian (NB), Bayesian networks (BN), lazy learning of Bayesian rules (LBR) and instance-based learner (IB1) are compared using a set of user profile data. According to our simulation results NB and IB1 classifiers have the highest classification accuracy with the lowest error rate. The obtained simulation results have been evaluated against the existing works of support vector machines (SVMs), decision trees (DTs) and neural networks (NNs).

Investigation of Mobile IPv6 and SIP integrated architectures for IMS and VoIP applications

Mobile IPv6 and SIP are protocols designed to support different types of mobility. Mobile IPv6 has been used to support mobility in IP networks and SIP has been used for voice over IP applications. It is the signalling protocol of the IP multimedia subsystem (IMS). In this paper both protocols have been simulated and compared in order to observe their performance for voice over IP (VoIP) applications. In this paper the architectures proposed by researchers in order to combine mobile IPv6 and SIP have also been investigated and compared to analyse their advantages and disadvantages. A network scenario, running mobile IPv6 and SIP for IMS, has also been simulated in order to evaluate the performance offered by the two protocols and to compare them with the results from the simulation of the pure mobile IPv6 and SIP architectures. The comparison shows that the combined scenario offers better performance similar to the one obtained using only mobile IPv6 with route optimization. The scenario simulated was also compared with the integrated architectures for mobile IPv6 and SIP that were investigated.

A Comparative Study of Selected Classification Accuracy in User Profiling

In recent years the used of personalization in service provisioning applications has been very popular. However, effective personalization cannot be achieved without accurate user profiles. A number of classification algorithms have been used to classify user related information to create accurate user profiles. In this study four different classification algorithms which are; naive Bayesian (NB), Bayesian Networks (BN), lazy learning of Bayesian rules (LBR) and instance-based learner (IB1) are compared using a set of user profile data. According to our simulation results NB and IB1 classifiers have the highest classification accuracy with the lowest error rate.

Computationally Significant Semantics in Pervasive Healthcare

Pervasive computing PerC is leading the way in a fast-growing trend of integrating transparently physical heterogeneous computational devices into our private and professional lives. The ubiquity of these devices and advances in developing software solutions in PerC across domains, have raised hopes for the creation of true wide-spread pervasive computing environments PCE. In this paper we explore the possibility of applying semantics of PCEs in the healthcare domain, and in Self Care Homes SeCH in particular, in order to define and comment on its computationally significant semantics. Our aim is to illustrate that we can manipulate the computationally significant semantics of SeCH through OWL/SWRL enabled ontologies, as candidate technologies for achieving effective and automated decision making in SeCH. The possibility of reasoning upon OWL/SWRL enabled concepts and creating computations from them, and enables the delivery of healthcare services to SeCH residents. They are automatically supported by software applications generated from the Assistive Self Care Systems ASeCS software architecture, which hosts our OWL/SWRL enabled ontology and its reasoning.

ASeCS: Assistive Self-Care Software Architectures for Delivering Service in Care Homes

We propose a layered and component based software architecture, which generates semantic software applications, for the purpose of delivering personalized services for residents in Self-Care Homes (SeCH). The architectural core layers accommodate software components which grasp and understand the semantic of various situations we may encounter in SeCH, through a variety of cyber-physical objects which co-exist in pervasive environments used in monitoring SeCH residents. The decision making on appropriate actions in SeCH is based on reasoning created by SWRL enabled OWL ontologies to ensure that in any situation, residents are delivered suitable and personalized healthcare services. The ASeCS architecture has been deployed through component based Java technologies, and uses OWL-API in order to seamlessly incorporate reasoning into software applications. ASeCS is SeCH specific, but provides a window of opportunities for creating modern and flexible software solutions for pervasive healthcare, where decision making solely depends on OWL/SWRL enabled computations.

Weighted Instance Based Learner (WIBL) for user profiling

With an increase in web-based products and services, user profiling has created opportunities for both businesses and other organizations to provide a channel for user awareness as well as to achieve high user satisfaction. Apart from traditional collaborative and content-based methods, a number of classification and clustering algorithms have been used for user profiling. Instance Based Learner (IBL) classifier is a comprehensive form of the Nearest Neighbour (NN) algorithm and it is suitable for user profiling as users with similar profiles are likely to share similar personal interests and preferences. In IBL every attribute has an equal effect on the classification regardless of their relevance. In this paper, we proposed a weighted classification method, namely Weighted Instance Based Learner (WIBL), to build and handle user profiles. With WIBL, we introduce Per Category Feature (PCF) method to IBL in order to distinguish the effect of attributes on classification. PCF is an attribute weighting method and it assigns weights to attributes using conditional probabilities. The direct use of this method with IBL is not possible. Hence, two possible solutions were also proposed to address this problem. This study is aimed to test the performance of WIBL for user profiling. To validate the performance of WIBL, a series of computer simulations were carried out. These simulations were conducted using a large user profile database that includes 5000 training and 1000 test instances (users). Here, each user is represented with three sets of profile information; demographic, interest and preference data. The results illustrate that WIBL with PCF methods performs better than IBL on user profiling by reducing the error up to 28% on the selected dataset.

Feature weighted clustering for user profiling

Personalization is the adaptation of the services to fit the user’s interests, characteristics and needs. The key to effective personalization is user profiling. Apart from traditional collaborative and content-based approaches, a number of classification and clustering algorithms have been used to classify user related information to create user profiles. However, they are not able to achieve accurate user profiles. In this paper, we present a new clustering algorithm, namely Multi-Dimensional Clustering (MDC), to determine user profiling. The MDC is a version of the Instance-Based Learner (IBL) algorithm that assigns weights to feature values and considers these weights for the clustering. Three feature weight methods are proposed for the MDC and, all three, have been tested and evaluated. Simulations were conducted with using two sets of user profile datasets, which are the training (includes 10,000 instances) and test (includes 1000 instances) datasets. These datasets reflect each user’s personal info...

ePerSpace: A Global Generic Network for Seamless Personalised Services

The ePerSpace project is developing and implementing a generic distributed networked system with wide ranging applications, accessible at home and globally anywhere else outside home. The project is creating an open, interoperable and trusted integration framework to create network enabled audiovisual systems and home platforms where home and personal devices can seamlessly work together providing personalised services, provisioning content adaptation, and managing service platforms. Using the personalisation information the system can recognise and form specific user communities towards which specific services can be directed. The paper presents the main concept and components of the system such as: the global service manager (GSeM) which handles service providers, users and the residential gateways (RGs), and local service manager (LSeM), which manages home and personal devices inside the home area network (HAN), and communication between GSeM and LSeM

Flexible channel allocation strategy for user service prioritisation in hierarchical cellular systems

A novel method for improving the spectrum efficiency of multi-layer cellular systems, by means of prioritisation based on the user mobility behaviour and service application has been proposed. The proposed system is shown to have a higher efficiency over a wider range of offered traffic, in comparison to the conventional FIFO queuing. Computational results suggest that using this new scheme, the system capacity for handover voice calls at 2% dropping rate can be improved by as much as 27% compared with FIFO. This benefit came at the expense of some increase in the dropping rate of the low priority calls. The improvement in capacity also reduces with service priority.