A recommender mechanism based on case-based reasoning

Case-based reasoning (CBR) algorithm is particularly suitable for solving ill-defined and unstructured decision-making problems in many different areas. The traditional CBR algorithm, however, is inappropriate to deal with complicated problems and therefore needs to be further revised. This study thus proposes a next-generation CBR (GCBR) model and algorithm. GCBR presents as a new problem-solving paradigm that is a case-based recommender mechanism for assisting decision making. GCBR can resolve decision-making problems by using hierarchical criteria architecture (HCA) problem representation which involves multiple decision objectives on each level of hierarchical, multiple-level decision criteria, thereby enables decision makers to identify problems more precisely. Additionally, the proposed GCBR can also provide decision makers with series of cases in support of these multiple decision-making stages. GCBR furthermore employs a genetic algorithm in its implementation in order to reduce the effort involved in case evaluation. This study found experimentally that using GCBR for making travel-planning recommendations involved approximately 80% effort than traditional CBR, and therefore concluded that GCBR should be the next generation of case-based reasoning algorithms and can be applied to actual case-based recommender mechanism implementation.     

Gaussian case-based reasoning for business failure prediction with empirical data in China

Case-based reasoning (CBR) is an easily understandable concept. Business failure prediction (BFP) is a valuable tool that can assist businesses take appropriate action when faced with the knowledge of the possibility of business failure. This study aims to improve the performance of a CBR system for BFP in terms of accuracy and reliability by constructing a new similarity measure, an area seldom researched in the domain of BFP. In order to fulfill this objective, we present a hybrid Gaussian CBR (GCBR) system and use it in BFP with empirical data in China. The heart of GCBR is similarity measure using Gaussian indicators. Feature distances between a pair of cases on each feature are transferred to Gaussian indicators by Gaussian transformations. A combiner is used to generate case similarity on the basis of the Gaussian indicators. Consensus of nearest neighbors is used to generate forecasting on the basis of case similarity. The new hybrid CBR system was empirically tested with data collected from the Shanghai Stock Exchange and Shenzhen Stock Exchange in China. We statistically validated our results by comparing them with multiple discriminant analysis, logistic regression, and two classical CBR systems. The results indicated that GCBR produces superior performance in short-term BFP of Chinese listed companies in terms of both predictive accuracy and coefficient of variation.     

Interactive Case-Based Reasoning in Sequential Diagnosis

Interactive trouble-shooting and customer help-desk support, both activities that involve sequential diagnosis, represent the majority of applications of case-based reasoning (CBR). An analysis is presented of the user-interface requirements of intelligent systems for sequential diagnosis. We argue that mixed-initiative dialogue, explanation of reasoning, and sensitivity analysis are essential to meet the needs of experienced as well as novice users. Other issues to be addressed by system designers include relevance and consistency in dialogue, tolerance of missing data, and timely provision of feedback to users. Many of these issues have previously been addressed by the developers of expert systems and the lessons learned may have important implications for CBR. We present a prototype environment for interactive CBR in sequential diagnosis, called CBR Strategist, which is designed to meet the identified requirements.     

Empirical evaluation of retrieval in case-based reasoning systemsusing modified cosine matching function

Case-based reasoning (CBR) supports ill-structured decision making by retrieving previous cases that are useful toward the solution of a new decision problem. The usefulness of previous cases is determined by assessing the similarity of a new case with the previous cases. In this paper, we present a modified form of the cosine matching function that makes it possible to contrast the two cases being matched and to include differences in the importance of features in the new case and the importance of features in the previous case. Our empirical evaluation of a CBR application to a diagnosis and repair task in an electromechanical domain shows that the proposed modified cosine matching function has a superior retrieval performance when compared to the performance of nearest-neighbor and the Tversky's contrast matching functions     

Real-time retrieval for case-based reasoning in interactive multiagent-based simulations

The aim of this paper is to present the principles and results about case-based reasoning adapted to real-time interactive simulations, more precisely concerning retrieval mechanisms. The article begins by introducing the constraints involved in interactive multiagent-based simulations. The second section presents a framework stemming from case-based reasoning by autonomous agents. Each agent uses a case base of local situations and, from this base, it can choose an action in order to interact with other autonomous agents or users’ avatars. We illustrate this framework with an example dedicated to the study of dynamic situations in football. We then go on to address the difficulties of conducting such simulations in real-time and propose a model for case and for case base. Using generic agents and adequate case base structure associated with a dedicated recall algorithm, we improve retrieval performance under time pressure compared to classic CBR techniques. We present some results relating to the performance of this solution. The article concludes by outlining future development of our project.     

New knowledge extraction technique using probability for case-based reasoning: application to medical diagnosis

Abstract: Case-based reasoning (CBR) has been used in various problem-solving areas such as financial forecasting, credit analysis and medical diagnosis. However, conventional CBR has the limitation that it has no criterion for choosing the nearest cases based on the probabilistic similarity of cases. It uses a fixed number of neighbors without considering an optimal number for each target case, so it does not guarantee optimal similar neighbors for various target cases. This leads to the weakness of lowering predictability due to deviation from desired similar neighbors. In this paper we suggest a new case extraction technique called statistical case-based reasoning. The main idea involves a dynamic adaptation of the optimal number of neighbors by considering the distribution of distances between potential similar neighbors for each target case. In order to do this, our technique finds the optimal distance threshold and selects similar neighbors satisfying the distance threshold criterion. We apply this new method to five real-life medical data sets and compare the results with those of the statistical method, logistic regression; we also compare the results with the learning methods C5.0, CART, neural networks and conventional CBR. The results of this paper show that the proposed technique outperforms those of many other methods, it overcomes the limitation of conventional CBR, and it provides improved classification accuracy .     

Two-step filtering datamining method integrating case-based reasoning and rule induction

Case-based reasoning (CBR) methods are applied to various target problems on the supposition that previous cases are sufficiently similar to current target problems, and the results of previous similar cases support the same result consistently. However, these assumptions are not applicable for some target cases. There are some target cases that have no sufficiently similar cases, or if they have, the results of these previous cases are inconsistent. That is, the appropriateness of CBR is different for each target case, even though they are problems in the same domain. Thus, applying CBR to whole datasets in a domain is not reasonable. This paper presents a new hybrid datamining technique called two-step filtering CBR and rule induction (TSFCR), which dynamically selects either CBR or RI for each target case, taking into consideration similarities and consistencies of previous cases. We apply this method to three medical diagnosis datasets and one credit analysis dataset in order to demonstrate that TSFCR outperforms the genuine CBR and RI.     

ASHSD-II: A computational model for litigation support

Abstract: In this paper a hybrid knowledge-based system which exploits both rule-based reasoning (RBR) and case-based reasoning (CBR) is presented. The issues of RBR and CBR in general in the context of legal knowledge-based systems and legislation in rule form and previously-decided cases in an interconnected graph form are discussed. It is possible for the user to select either reasoning method (RBR or CBR), or indicate no preference. The rule base of this system consists of two types of rule. The first type of rule determines which options are legally applicable. The second type indicates how the courts are likely to act within the range of options available, which is determined by the first type of rule. When CBR is selected, the system uses the features of previously-decided cases to select the most similar cases to the situation that is described in the input and displays their details of decisions. In case of the selection of no preference option, the system applies RBR and CBR method separately, and then presents results based on an automated relative rating of the qualities of the RBR (based on the second type of rules) and CBR advice. These ideas have been implemented in a prototype system, known as A dvisory S upport for H ome S ettlement in D ivorce (ASHSD-II).     

Hybrid genetic algorithms and case-based reasoning systems for customer classification

Abstract: Because of its convenience and strength in complex problem solving, case-based reasoning (CBR) has been widely used in various areas. One of these areas is customer classification, which classifies customers into either purchasing or non-purchasing groups. Nonetheless, compared to other machine learning techniques, CBR has been criticized because of its low prediction accuracy. Generally, in order to obtain successful results from CBR, effective retrieval of useful prior cases for the given problem is essential. However, designing a good matching and retrieval mechanism for CBR systems is still a controversial research issue. Most previous studies have tried to optimize the weights of the features or the selection process of appropriate instances. But these approaches have been performed independently until now. Simultaneous optimization of these components may lead to better performance than naive models. In particular, there have been few attempts to simultaneously optimize the weights of the features and the selection of instances for CBR. Here we suggest a simultaneous optimization model of these components using a genetic algorithm. To validate the usefulness of our approach, we apply it to two real-world cases for customer classification. Experimental results show that simultaneously optimized CBR may improve the classification accuracy and outperform various optimized models of CBR as well as other classification models including logistic regression, multiple discriminant analysis, artificial neural networks and support vector machines.     

DSS for computer security incident response applying CBR and collaborative response

Recently, as hacking attempts increase dramatically; most enterprises are forced to employ some safeguards for hacking proof. For example, firewall or IPS (Intrusion Prevention System) selectively accepts the incoming packets, and IDS (Intrusion Detection System) detects the attack attempts from network. The latest version of firewall works in cooperation with IDS to immediately response to hacking attempts. However, it may make false alarms that misjudge normal traffic as hacking traffic and cause network problems to block the normal IP address by false alarms. By these false alarms made by IDS, system administrators or CSOs make wrong decisions and important data may be exposed or the availability of network or server system may be exhausted. Therefore, it is important to minimize the false alarms.As a way of minimizing false alarms and supporting adequate decisions, we suggest the RFM (Recency, Frequency, Monetary) analysis methodology, which analyzes log files with incorporating three criteria of recency, frequency and monetary with statistical process control chart, and thus leads to an intuitive detection of anomaly and misuse events. Moreover, to cope with hacking attempts proactively, we apply CBR (case based reasoning) to find out similarities between already known hacking patterns and new hacking patterns. With the RFM analysis methodology and CBR, we develop DSS which can minimize false alarms and decrease the time to respond to hacking events. In case that RFM analysis module finds out unknown viruses or worms occurred, this CBR system matches the most similar incident case from case-based database. System administrators can easily get information about how to fix and how we fixed in similar cases. And CSOs can build a blacklist of frequently detected IP addresses and users. This blacklist can be used for incident handling.Finally, we propose collaborative incident response system with DSS, this distributed agent systems interactively exchange the suspicious users and source IP addresses data and decide who is true-anomalous users and which IP addresses is the most riskiest and then deny all connections from that users and IP addresses automatically with less false-positives.     

基于事例的推理(CBR)研究综述

基于事例的推理(CBR)作为一种增量式的学习方法,规避了传统人工智能在知识获取上的瓶颈问题,逐渐引起人工智能领域研究者的关注。对基于事例的推理(CBR)现有研究工作进行逻辑上的梳理和系统的总结,有助于今后研究工作的开展,具有深远的理论意义。该文首次提出基于事例的推理(CBR)研究的逻辑体系结构,并在此逻辑分析的基础上,从基本理论、关键技术和实践应用三方面进行了综述,对其中关键、通用的方法和技术进行了比较和评价。最后,对未来的研究方向进行了展望。     

Applying RBR and CBR to develop a VR based integrated system for machining fixture design

A fixture is a special tool used to accurately and stably locate the workpiece during machining process. Proper fixture design improves the quality and production of parts and also facilitates the interchangeability of parts that is prevalent in much of modern manufacturing. This study combines the rule-based reasoning (RBR) and case-based reasoning (CBR) method for machining fixture design in a VR based integrated system. In this paper, an approach combines the RBR and fuzzy comprehensive judgment method is proposed for reasoning suitable locating schemes and locating features. Based on the reasoning results, a CBR method for machining fixture design is then presented. This method could help designers, by referencing previous design cases, to make a conceptual fixturing solution quickly. Finally, the implementation of proposed system is outlined and cases study has been used to demonstrate the applicability of the proposed approach.     

Prototype-based management of business process exception cases

Business process optimization may require to deviate from a default process model, in response to unexpected situations, thus raising exceptions. In this paper, we present a system for supporting end users in handling exceptions in business process management, which exploits the case-based reasoning (CBR) methodology. CBR offers the advantage of relying on operative knowledge, thus reducing the cost of knowledge elicitation, with respect to other methodologies.     

Improving user experience with case-based reasoning systems using text mining and Web 2.0

Many CBR systems have been developed in the past. However, currently many CBR systems are facing a sustainability issue such as outdated cases and stagnant case growth. Some CBR systems have fallen into disuse due to the lack of new cases, case update, user participation and user engagement. To encourage the use of CBR systems and give users better experience, CBR system developers need to come up with new ways to add new features and values to the CBR systems. The author proposes a framework to use text mining and Web 2.0 technologies to improve and enhance CBR systems for providing better user experience. Two case studies were conducted to evaluate the usefulness of text mining techniques and Web 2.0 technologies for enhancing a large scale CBR system. The results suggest that text mining and Web 2.0 are promising ways to bring additional values to CBR and they should be incorporated into the CBR design and development process for the benefit of CBR users.     

基于微粒群的案例推理方法研究

距离测度是案例检索的关键问题,它直接影响案例检索精度.针对距离测度进行研究,提出一种基于微粒群方法的自学习距离测度,并将该自学习距离测度引入案例推理中,使案例推理在处理由相关属性表述的案例时有了合理的解决方法,从而扩展了案例推理的应用范围.最后,利用实际数据与UCI数据对基于新距离测度的案例推理技术进行了仿真实验,实验结果表明,与其他方法相比,该方法可以提高案例检索的准确性.     

Incremental development of CBR strategies for computing project cost probabilities

This paper presents three CBR systems that have been developed over seven years in collaboration with two industrial partners. In this research, case based reasoning (CBR) is used to compute costs of construction projects. In contrast with previous work in the field of CBR, the focus is on choosing strategies that are compatible with user needs and characteristics. Comparing the three strategies reveals advantages and drawbacks while illustrating a “real-life” evolution of a CBR architecture in an industrial context. An important conclusion is that the ways users perform tasks have a direct influence on the best architecture for the CBR system (e.g. transformational/derivational analogy). Incremental development of strategies in the final system improves user interaction, expedites time consuming tasks and favours identification of synergy between techniques such as CBR and data mining.     

基于人机协同的潜在意图检测模型

潜在意图检测旨在通过意图主体行为推理意图主体的隐式意图,从而在更高的层面理解意图主体潜在的真实意图.提出了一种多领域数据环境下人机协同的潜在意图检测模型和技术框架.该意图检测模型扩展了动态意图表示形式DIS,能够适应多领域数据和交互式推理的意图表示需要.通过定义交互原语,确定了人机协同交互的语言规范.通过功能框架的设计,提供了潜在意图检测的技术实现途径.     

Application of a 3NN+1 based CBR system to segmentation of the notebook computers market

Case-based reasoning system (CBR) has been widely applied to the issue of market segmentation. Most of previous studies focused on dividing customers into two groups. Consequently, traditional voting method used for two groups in CBR would become inappropriate when one would like to divide customers into three groups through some segmentation variable. In this paper, a new voting method called 3NN+1 is proposed to bridge the gap. To make the inference of the 3NN+1 based CBR system more efficient, the features and instances (or cases) for reasoning is selected simultaneously by means of genetic algorithms. This new system is applied to a real case of notebook market to demonstrate its usefulness for market segmentation. From the results of the real case, it shows that the system would be valuable to enterprises, when dividing customers into three groups in compliance with their purchasing behaviors for developing marketing strategies.     

案例推理变权值引擎模型及权值计算方法

在案例推理（CBR）案例检索匹配中,不同案例通常由不同的特征构成。而传统的CBR引擎模型大多采用固定权值模式,导致系统在匹配精度方面的性能很低。为了解决这一问题,提出一种CBR变权值引擎模型,在其特征权值计算模块引入人机互动机制,基于群决策法计算主观权值,提出依据专家个体和群体决策差异的主观权值调整方法;基于相似粗糙集法计算客观权值。最后设计了一种综合权值调整算法,通过计算主观权值和客观权值间的距离,判断两者的偏离程度,从而推导出权值调整系数,得到最终的权值调整结果。通过网络攻击案例进行的算例分析和仿真实验验证了上述方法的正确性和优越性。