Search

A key application of wearable sensors is remote patient monitoring, which facilitates clinicians to observe patients non-invasively, by examining the time series of sensor readings. For analysis of such time series, a recently proposed technique is Matrix Profile (MP). While being effective for certain time series mining tasks, MP depends on a key input parameter, the length of subsequences for which to search. We demonstrate that MP’s dependency on this input parameter impacts its effectiveness for finding patterns of interest. We focus on finding consecutive repeating patterns (CRPs), which represent human activities and exercises whilst tracked using wearable sensors. We demonstrate that MP cannot detect CRPs effectively and extend it by adding a locality preserving index. Our method automates the use of MP, and reduces the need for data labeling by experts. We demonstrate our algorithm’s effectiveness in detecting regions of CRPs through a number of real and synthetic datasets.

If effort estimates are not easily assessed upfront by software maintainers we may have serious problems with large maintenance projects, or when we make repeated maintenance changes to software. This is particularly problematic when inaccurate estimates of the required resources leads to serious negotiation issues. The development of a Categorisation of Maintenance Effort (COME) matrix enables an overall summary of software maintenance changes and maintenance effort to be shown, upfront, to software practitioners. This can occur without any consideration or use of other effort estimation techniques whose results, when used to estimate effort, can appear complicated and it may not be clear how accurate their estimates may be. We use a simple approach to categorizing maintenance effort data using five steps. We use regression analysis with Jorgensen's 81 datasets to evaluate the selected variables to find out the true efficacy of our approach: 1) adaptive changes and functional changes with maintenance effort predicted from low to high, 2) high predicted effort when updating KSLOC for software maintenance changes, 3) find that more lines of source codes do not imply that more software maintenance effort is needed, 4) find no significant relationship when we consider the age of the application and 5) find that at least 20 application sizes between KSLOC of 100, 200, 400 and 500 have a low predicted software maintenance effort. Our experiment shows that using the COME matrix is an alternative approach to other cost estimation techniques for estimating effort for repeated requirement changes in large software maintenance projects.

In attempting to read a scholarly article, learners (students) often struggle with problems of comprehension. It is likely that when a scholar writes a paper and discusses a new idea or a method within a particular discipline, they assume that readers have a scholarly background enabling them to understand the paper content in detail. Such an assumption is not justified, and there is evidence that students who are learning to carry out research for the first time find that understanding scholarsâY papers is not always an easy task. Deciding which RE (Requirement Elicitation) technique and tool to apply to issues can become complicated for educators constructing a learning approach which emphasises that group learners should read scholarly articles in a short time and be able to summarise the content from the article, while at the same time allowing learners to devise solutions that will, enhance their creativity and innovation, allowing them to explore how an idea in a paper could be integrated into an industry application. A case study in this paper introduces a university framework, which can be applied to aid students in their decision-making on selecting a presentation technique; that is, choosing role-playing as the appropriate technique for an effective learning outcome.

We propose that the key to building intelligent negotiating agents is to take an agents historic observations as primitive, to model that agents changing uncertainty in that information, and to use that model as the foundation for the agents reasoning. We describe an agent architecture, with an attendant theory, that is based on that model. In this approach, the utility of contracts, and the trust and reliability of a trading partner are intermediate concepts that an agent may estimate from its information model. This enables us to describe intelligent agents that are not necessarily utility optimisers, that value information as a commodity, and that build relationships with other agents through the trusted exchange of information as well as contracts.

With advancements in technology, personal computing devices are better adapted for and further integrated into people's lives and homes. The integration of technology into society also results in an increasing desire to control who and what has access to sensitive information, especially for vulnerable people including children and the elderly. With blockchain coming in to the picture as a technology that can revolutionise the world, it is now possible to have an immutable audit trail of locational data over time. By controlling the process through inexpensive equipment in the home, it is possible to control whom has access to such personal data. This paper presents a blockchain based family security system for tracking the location of consenting family members' smart phones. The locations of the family members' smart phones are logged and stored in a private blockchain which can be accessed through a node installed in the family home on a computer. The data for the whereabouts of family members stays within the family unit and does not go to any third party. The system is implemented in a small scale (one miner and two other nodes) and the technical feasibility is discussed along with the limitations of the system. Further research will cover the integration of the system into a smart home environment, and ethical implementations of tracking, especially of vulnerable people, using the immutability of blockchain.

Educational theory has purported the notion that student-centric modes of learning are more effective in enhancing student engagement and by extension, learning outcomes. However, the translation of this theoretical pedagogy of learning into an applied model for medical training has been fraught with difficulty due to the structural complexity of creating a classroom environment that enables students to exercise full autonomy. In this paper, we propose an intelligent computational e-learning platform for case-based learning (CBL) in Medicine that enriches and enhances the learning experiences of medical students by exposing them to simulated real-world clinical contexts. We argue that computational systems in Medicine should not merely provide a passive outlay of information, but instead promote active engagement through an immersive learning experience. This is achieved through a digital platform that renders a virtual patient simulation, which allows students to assess, diagnose, treat and test patients as they would in the real-world.

Web monitoring systems and meta-search engines were designed to provide time and coverage critical services, where time critical means that new information should be provided as soon as it is publicized on the web and coverage critical means that any information should not be missed by the systems. We have analyzed coverage and timeliness of three commercial search engines with the web page monitoring results to investigate how rapidly and how efficiently web monitoring system and meta-search engines collect and provide newly published web information. We have also assessed how the meta-search engines might improve coverage and timeliness by providing collective services. Our experiment results show that commercial search engines still cover 65% ~ 75% of newly published information, taking from five to 13 days to retrieve the information. Theoretically, meta-search engines discover up to 86% of all published data and shorten delay time up to 8 days.

This paper describes and compares the use of methods based on Ngrams(specifically trigrams and pentagrams), together with five features, to recognise the syntactic and semantic categories of numeral strings representing money, number, date, etc., in texts. The system employs three interpretation processes: word N-grams construction with a tokeniser; rule-based processing of numeral strings; and N-gram-based classification. We extracted numeral strings from 1,111 online newspaper articles. For numeral strings interpretation, we chose 112 (10%) of 1,111 articles to provide unseen test data (1,278 numeral strings), and used the remaining 999 articles to provide 11,525 numeral strings for use in extracting N-gram-based constraints to disambiguate meanings of the numeral strings. The word trigrams method resulted in 83.8% precision, 81.2% recall ratio, and 82.5% in F-measurement ratio. The word pentagrams method resulted in 86.6% precision, 82.9% recall ratio, and 84.7% in F-measurement ratio.

t is necessary to provide a method to store Web information effectively so it can be utilised as a future knowledge resource. A commonly adopted approach is to classify the retrieved information based on its content. A technique that has been found to be suitable for this purpose is Multiple Classification Ripple-Down Rules (MCRDR). The MCRDR system constructs a classification knowledge base over time using an incremental learning process. This incremental method of acquiring classification knowledge suits the nature of Web information because it is constantly evolving and being updated. However, despite this advantage, the classification knowledge of the MCRDR system is not often utilised for browsing the classified information. This is because it does not directly organise the knowledge in a way that is suitable for browsing. As a result, often an alternate structure is utilised for browsing the information which is usually based on a user's abstract understanding of the information domain. This study investigated the feasibility of utilising the classification knowledge acquired through the use of the MCRDR system as a resource for browsing information retrieved from the WWW. A system was implemented that used the concept lattice-based browsing scheme of Formal Concept Analysis (FCA) to support the browsing of documents based on the MCRDR classification knowledge. The feasibility of utilising classification knowledge as a resource for browsing documents was evaluated statistically. This was achieved by comparing the concept lattice-based browsing approach to a standard one that utilises abstract knowledge of a domain as a resource for browsing the same documents.

Real-world document classification is an open-ended problem, rather than a close-ended problem, because the document classification domain continually evolves as the time passes. Unlike the close-ended document classification, the participants in the open-ended problem actively take part in the problem solving process. For this reason, it is important to understand the problem solver's behavioral characteristics. This paper proposes a thorough analysis of them. We found that the problem solving strategies are significantly different among participants because of individual differences in cognition among participants

Web services are key applications in business-to-business, business-to-customer, and enterprise applications integration solutions. As the mobile Internet becomes one of the main methods for information delivery, mobile Web Services are regarded as a critical aspect of e-business architecture. In this paper, we proposed a mobile Web Services middleware that converts conventional Internet services into mobile Web services. We implemented a WSDL (Web Service Description Language) builder that converts HTML/XML into WSDL and a SAOP (Simple Object Access Protocol) message processor. The former minimizes the overhead cost of rebuilding mobile Web Services and enables seamless services between wired and wireless Internet services. The latter enhances SOAP processing performance by eliminating the Servlet container (Tomcat), a required component of typical Web services implementation. Our system can completely support standard Web Services protocol, minimizing communication overhead, message processing time, and server overload. Finally we compare our empirical results with those of typical Web Services

This paper focuses on the development of a maintainable information filtering system. The simple and efficient solution to this problem is to block the Web sites by URL, including IP address. However, it is not efficient for unknown Web sites and it is difficult to obtain complete block list. Content based filtering is suggested to overcome this problem as an additional strategy of URL filtering. The manual rule based method is widely applied in current content filtering systems, but they overlook the knowledge acquisition bottleneck problems. To solve this problem, we employed the Multiple Classification Ripple-Down Rules (MCRDR) knowledge acquisition method, which allows the domain expert to maintain the knowledge base without the help of knowledge engineers. Throughout this study, we will prove the MCRDR based information filtering system can easily prevent unknown Web information from being delivered and easily maintain the knowledge base for the filtering system.

These days, billions of Web pages are created with HTML or other markup languages. They only have a few uniform structures and contain various authoring styles compared to traditional text-based documents. However, users usually focus on a particular section of the page that presents the most relevant information to their interest. Therefore, Web documents classification needs to group and filter the pages based on their contents and relevant information. Many researches on Web mining report on mining Web structure and extracting information from web contents. However, they have focused on detecting tables that convey specific data, not the tables that are used as a mechanism for structuring the layout of Web pages. Case modeling of tables can be constructed based on structure abstraction. Furthermore, Ripple Down Rules (RDR) is used to implement knowledge organization and construction, because it supports a simple rule maintenance based on case and local validation.

This paper focuses on real world Web document classification problem. Real world Web documents classification has different problems compare to experimental based classification. Web documents have been continually increased and their themes also have been continually changed. Furthermore, domain users' knowledge is not fixed apart from classification environments. They learn from classification experience, broaden their knowledge, and tend to reclassify pre-classified Web documents according to newly obtained knowledge to fit various contexts. To handle these kinds of problems, we use Multiple Classification Ripple-Down Rules (MCRDR) knowledge acquisition method. The MCRDR based document classification enables domain users to elicit their domain knowledge incrementally and revise their knowledge base (KB), and consequently reclassify preclassified documents according to context changes. Our experiment results show MCRDR document classifier performs these tasks successfully in the real world.

Noise data in the Web document significantly affect on the performance of the Web information management system. Many researchers have proposed document structure based noise data elimination methods. In this paper, we propose a different approach that uses a redundant information elimination approach in the Web documents from the same URL path. We propose a redundant word/phrase filtering method for single or multiple tokenizations. We conducted two experiments to examine efficiency and effectiveness of our filtering approaches. Experimental results show that our approach produces a high performance in these two criteria

The classical concept of a finite state automaton (FSA) has long been used in computer science to classify formal languages regarded as sets of strings. This paper surveys and unifies recent results on strings recognized by FSA defined by labeled directed graphs, and highlights several interesting open questions. In particular, one of our open questions deals with a new method of applying a related construction to classification

Real-world document classification is an open-ended problem, rather than a close-ended problem, because the document classification domain continually evolves as the time passes. Unlike the close-ended document classification, the participants in the open-ended problem actively take part in the problem solving process. For this reason, it is important to understand the problem solver's behavioral characteristics. This paper proposes a thorough analysis of them. We found that the problem solving strategies are significantly different among participants because of individual differences in cognition among participants

t is necessary to provide a method to store Web information effectively so it can be utilised as a future knowledge resource. A commonly adopted approach is to classify the retrieved information based on its content. A technique that has been found to be suitable for this purpose is Multiple Classification Ripple-Down Rules (MCRDR). The MCRDR system constructs a classification knowledge base over time using an incremental learning process. This incremental method of acquiring classification knowledge suits the nature of Web information because it is constantly evolving and being updated. However, despite this advantage, the classification knowledge of the MCRDR system is not often utilised for browsing the classified information. This is because it does not directly organise the knowledge in a way that is suitable for browsing. As a result, often an alternate structure is utilised for browsing the information which is usually based on a user's abstract understanding of the information domain. This study investigated the feasibility of utilising the classification knowledge acquired through the use of the MCRDR system as a resource for browsing information retrieved from the WWW. A system was implemented that used the concept lattice-based browsing scheme of Formal Concept Analysis (FCA) to support the browsing of documents based on the MCRDR classification knowledge. The feasibility of utilising classification knowledge as a resource for browsing documents was evaluated statistically. This was achieved by comparing the concept lattice-based browsing approach to a standard one that utilises abstract knowledge of a domain as a resource for browsing the same documents.

Web monitoring systems and meta-search engines were designed to provide time and coverage critical services, where time critical means that new information should be provided as soon as it is publicized on the web and coverage critical means that any information should not be missed by the systems. We have analyzed coverage and timeliness of three commercial search engines with the web page monitoring results to investigate how rapidly and how efficiently web monitoring system and meta-search engines collect and provide newly published web information. We have also assessed how the meta-search engines might improve coverage and timeliness by providing collective services. Our experiment results show that commercial search engines still cover 65% ~ 75% of newly published information, taking from five to 13 days to retrieve the information. Theoretically, meta-search engines discover up to 86% of all published data and shorten delay time up to 8 days.

Web services are key applications in business-to-business, business-to-customer, and enterprise applications integration solutions. As the mobile Internet becomes one of the main methods for information delivery, mobile Web Services are regarded as a critical aspect of e-business architecture. In this paper, we proposed a mobile Web Services middleware that converts conventional Internet services into mobile Web services. We implemented a WSDL (Web Service Description Language) builder that converts HTML/XML into WSDL and a SAOP (Simple Object Access Protocol) message processor. The former minimizes the overhead cost of rebuilding mobile Web Services and enables seamless services between wired and wireless Internet services. The latter enhances SOAP processing performance by eliminating the Servlet container (Tomcat), a required component of typical Web services implementation. Our system can completely support standard Web Services protocol, minimizing communication overhead, message processing time, and server overload. Finally we compare our empirical results with those of typical Web Services

In general, document classification research focuses on the automated placement of unseen documents into pre-defined categories. This is regarded as one core technical component of knowledge management systems, because it can support to handle explicit knowledge more systematically and improve knowledge sharing among the users. Document classification in knowledge management systems should support incremental knowledge acquisition and maintenance because of the dynamic knowledge changes involved. We propose the MCRDR document classifier as an incremental and maintainable document classification solution. Even though our system successfully supported personal level document classification, we did not examine its capability as a document classification tool in multi-user based knowledge management contexts. This paper focuses on the analysis of document classification results performed by multiple users. Our analysis reveals that even though the same documents and the classification structure are given to the users, they have very different document classification patterns and different acceptance results for each other’s classification results. Furthermore, our results show that the integration of multiple users’ classification may improve document classification performance in the knowledge management context.

Background and objective: The aim was to develop and evaluate a pilot version of a knowledge-based system that can identify existing and potential medication-related problems from patient information. This intelligent system could directly support pharmacists and other health professionals providing medication reviews. Methods: Rather than being based on static rules to trigger alerts, this system utilizes a multiple classification ripple-down rules approach, which allows the user to build rules incrementally and improve the accuracy of the knowledge base in identifying medication-related problems while the system is in use, with no outside assistance or training. The system contextualizes the potential drug therapy problems by taking into consideration the patient's demographics, and other medical condition and drugs. The system is capable of both being instructed in the domain of medication review through its routine use by an expert, and acting similarly to the expert when analysing genuine medication review cases. The system was handed over to an experienced clinical pharmacist (expert), with no knowledge or conclusions preloaded into the system. The expert was then able to add the case details and generate the rules required for 126 actual medication review cases. Results: Over 250 rules were generated from the review cases, incorporating demographics, medical history, symptoms, medications and pathology results from these cases. At the completion of the cases, more than 80% of the potential medication-related problems identified by the expert were also detected by the system. The false positive rate, or number of incorrect medication-related problems identified by the system, was <10% overall and was zero for the last 15 cases analysed. The system found significantly more potential medication-related problems than the expert, with the system consistently remaining at least one finding ahead. There was a high incidence of missed potential medication-related pro

Web information publishers are interested in how well their information is indexed by major search engines within short time, because the search engines are a main information access point for their web sites. Being the crawling policy of each search engine is usually commercial secret, it is useful to estimate each search engine’s coverage and delay with known predicator variables. This paper proposes forecasting models for service coverage and delay of search engines in the Australian government area using predictor variables, identified from the crawling policies of academic papers, and statistical regression methods. The Logistic regression method was employed for coverage forecast and Poisson regression method for delay. Our research results show that different explanatory variables were chosen for constructing models and their importance significantly varies among search engines.

A method is proposed for the representation of localised features using disjoint sub-images taken from several datasets of retinal images for use within an incremental learning system. A tile-based localised adaptive threshold selection method was taken for vessel segmentation based on separate colour components. Arteriole-venous differentiation was done using the composite of these components and high quality fundal images. Vessel segmentation performance was evaluated on the DRIVE and STARE datasets achieving average specificity of 0.9379 and sensitivity of 0.5924.

This paper describes and compares the use of methods based on Ngrams(specifically trigrams and pentagrams), together with five features, to recognise the syntactic and semantic categories of numeral strings representing money, number, date, etc., in texts. The system employs three interpretation processes: word N-grams construction with a tokeniser; rule-based processing of numeral strings; and N-gram-based classification. We extracted numeral strings from 1,111 online newspaper articles. For numeral strings interpretation, we chose 112 (10%) of 1,111 articles to provide unseen test data (1,278 numeral strings), and used the remaining 999 articles to provide 11,525 numeral strings for use in extracting N-gram-based constraints to disambiguate meanings of the numeral strings. The word trigrams method resulted in 83.8% precision, 81.2% recall ratio, and 82.5% in F-measurement ratio. The word pentagrams method resulted in 86.6% precision, 82.9% recall ratio, and 84.7% in F-measurement ratio.

The Four Legged League is a division of the RoboCup initiative that uses Sony AIBOTM robots to further robotics research. Most participants implement vision systems that use the color of objects to perform identification. Calibration of the color classification system must be done and any changes to the lighting of the environment after calibration reduces the accuracy of the system, often to a point at which the robot is effectively blind. This study investigates the relationships in the color data of image pixels between lighting conditions in an effort to identify trends that can be used as the basis of a rule-based system. The aim of the system is to identify the current lighting level as one of a set of known conditions. The proposed systems uses the color data of image pixels and information about the AIBO’s location and orientation to identify lighting levels, allowing a vision system to switch to an appropriate pre-configured calibration

Multiple Classification Ripple Down Rules (MCRDR) is a simple and effective knowledge acquisition technique that produces representations, or knowledge maps, of a human experts' knowledge of a particular domain. This knowledge map can then be used to automate and help the user perform classification and categorisation of cases while still being able to add more refined knowledge incrementally. While MCRDR has been applied in many domains, work on understanding the meta-knowledge acquired or using the knowledge to derive new information is still in its infancy. This paper will introduce a technique called Rated MCRDR (RM), which looks at deriving and learning information about both linear and non-linear relationships between the multiple classifications within MCRDR. This method uses the knowledge received in the MCRDR knowledge map to derive additional information that allows improvements in functionality within existing domains, to which MCRDR is currently applied, as well as opening up the possibility of new problem domains. Preliminary testing shows that there exists a strong potential for RM to quickly and effectively learn meaningful ratings.

This paper describes and compares the use of methods based on N-grams (specifically trigrams and pentagrams), together with five features, to recognise the syntactic and semantic categories of numeral strings representing money, number, date, etc., in texts. The system employs three interpretation processes: word N-grams construction with a tokeniser; rule-based processing of numeral strings; and N-gram-based classification. We extracted numeral strings from 1,111 online newspaper articles. For numeral strings interpretation, we chose 112 (10%) of 1,111 articles to provide unseen test data (1,278 numeral strings), and used the remaining 999 articles to provide 11,525 numeral strings for use in extracting N-gram-based constraints to disambiguate meanings of the numeral strings. The word trigrams method resulted in 83.8% precision, 81.2% recall ratio, and 82.5% in F-measurement ratio. The word pentagrams method resulted in 86.6% precision, 82.9% recall ratio, and 84.7% in F-measurement ratio.

Multiple Classification Ripple Down Rules (MCRDR) is a knowledge acquisition technique that produces representations, or knowledge maps, of a human expert's knowledge of a particular domain. However, work on gaining an understanding of the knowledge acquired at a deeper meta-level or using the knowledge to derive new information is still in its infancy. This paper will introduce a technique called Weighted MCRDR (WM), which looks at deriving and learning information about the relationships between multiple classifications within MCRDR by calculating a meaningful rating for the task at hand. This is not intended to reduce the knowledge acquisition effort for the expert. Rather, it is attempting to use the knowledge received in the MCRDR knowledge map to derive additional information that can allow improvements in functionality of MCRDR in many problem domains. Preliminary testing shows that there exists a strong potential for WM to quickly and effectively learn meaningful weightings.

The focus of traditional conversational agents is placed on natural language processing and understanding the needs of the user. These agents are typically implemented for specific domains such that domain knowledge and conversations are built manually. Domain knowledge of these agents are encoded as conversational content which causes problems is magnified by the lack of knowledge acquisition tools and, as a result, agents find it difficult to adapt to different domains and to update existing knowledge bases. The framework proposed in this paper aims to rectify this problem by building a module to handle knowledge acquisition. The module acquires knowledge through a case-based methodology called Ripple Down Rules (RDR); a technique that has been employed successfully across a host of expert systems

This paper introduces an augmentation hybrid system, referred to as Rated MCRDR. It uses Multiple Classification Ripple Down Rules (MCRDR), a simple and effective knowledge acquisition technique, combined with a neural network.

Knowledge Based Systems (KBS) have long wrestled with the problem of incomplete knowledge that occasionally causes them to make ridiculous conclusions. Knowledge engineers have searched for methodologies that allow for less brittle systems. Additionally, KBS systems for general knowledge have been developed to try and build background information that a system can fall back on when they cannot find a conclusion in their specific domain. However, it is next to impossible to include all the required knowledge to completely eradicate the inherent brittleness of these systems. This paper presents a method for predicting when a case being presented to the KBS is outside its current knowledge. When the system notices such a case it provides a warning allowing the user to investigate the case further. This preliminary study of the system has been tested using a simulated expert with randomly generated data sets. It shows that this system has great potential for predicting almost every error and rarely issuing warnings for correct conclusions. Such a system could significantly reduce the knowledge acquisition task for an expert. These results clearly show the potential of such a system and that further investigation with recognised data sets and real user tests should be performed.

This paper focuses on the development of a maintainable information filtering system. The simple and efficient solution to this problem is to block the Web sites by URL, including IP address. However, it is not efficient for unknown Web sites and it is difficult to obtain complete block list. Content based filtering is suggested to overcome this problem as an additional strategy of URL filtering. The manual rule based method is widely applied in current content filtering systems, but they overlook the knowledge acquisition bottleneck problems. To solve this problem, we employed the Multiple Classification Ripple-Down Rules (MCRDR) knowledge acquisition method, which allows the domain expert to maintain the knowledge base without the help of knowledge engineers. Throughout this study, we will prove the MCRDR based information filtering system can easily prevent unknown Web information from being delivered and easily maintain the knowledge base for the filtering system.

Expert Systems in data mining generally use knowledge extraction methods to form a classifier or predictor. These have the advantage of forming high quality results due to the inclusion of expert knowledge. The problem, however, is that they do not allow for autonomous knowledge discovery. Therefore, such systems will only find results that the expert is capable of giving examples about and will not find unknown patterns. Thus, Knowledge Discovery in Database (KDD) generally relies on other machine learning tools, forgoing the advantage of expert knowledge. This paper presents a method for incrementally building a knowledge base using expert knowledge in such a way that potentially the system is still able to autonomously discover unknown patterns. It does this through prudence checking of the knowledge base for each case presented and informs the expert when the knowledge base has an inconsistency needing clarification. Initial results show strong potential for the system in identifying misclassifications allowing its potential application to KDD.

The article describes two new clustering algorithms for DNA nucleotide sequences, summarizes the results of experimental analysis of performance of these algorithms for an ITS-sequence data set, and compares the results with known biologically significant clusters of this data set. It is shown that both algorithms are efficient and can be used in practice.