Showing papers in "Knowledge Based Systems in 2001"

A resource limited artificial immune system for data analysis

Abstract: This paper presents a resource limited artificial immune system (RLAIS) for data analysis. The work presented here builds upon previous work on artificial immune systems (AIS) for data analysis. A population control mechanism, inspired by the natural immune system, has been introduced to control population growth and allow termination of the learning algorithm. The new algorithm is presented, along with the immunological metaphors used as inspiration. Results are presented for Fisher Iris data set, where very successful results are obtained in identifying clusters within the data set. It is argued that this new resource-based mechanism is a large step forward in making AISs a viable contender for effective unsupervised machine learning and allows for not just a one shot learning mechanism, but a continual learning model to be developed.

Information access in context

Abstract: Our central claim is that user interactions with productivity applications (e.g. word processors, Web browsers, etc.) provide rich contextual information that can be leveraged to support just-in-time access to task-relevant information. As evidence for our claim, we present Watson, a system which gathers contextual information in the form of the text of the document the user is manipulating, in order to proactively retrieve documents from distributed information repositories related to task at hand, as well as process explicit requests in the context of this task. We close by describing the results of several experiments with Watson, which show it consistently provides useful information to its users. The experiments also suggest that, contrary to the assumptions of many system designers, similar documents are not necessarily useful documents in the context of a particular task.

A data mining approach to the prediction of corporate failure

Abstract: This paper uses a data mining approach to the prediction of corporate failure. Initially, we use four single classifiers — discriminant analysis, logistic regression, neural networks and C5.0 — each based on two feature selection methods for predicting corporate failure. Of the two feature selection methods — human judgement based on financial theory and ANOVA statistical method — we found the ANOVA method performs better than the human judgement method in all classifiers except discriminant analysis. Among the individual classifiers, decision trees and neural networks were found to provide better results. Finally, a hybrid method that combines the best features of several classification models is developed to increase the prediction performance. The empirical tests show that such a hybrid method produces higher prediction accuracy than individual classifiers.

Making decisions: using Bayesian nets and MCDA

Abstract: Bayesian belief nets (BBNs) have proven to be an extremely powerful technique for reasoning under uncertainty. We have used them in a range of real applications concerned with predicting properties of critical systems. In most of these applications we are interested in a single attribute of the system such as safety or reliability. Although such BBNs provide important support for decision making, in many circumstances we need to make decisions based on multiple criteria. For example, a BBN for predicting the safety of a critical system cannot be used to make a decision about whether or not the system should be deployed. This is because such a decision must be based on criteria other than just safety (cost, politics, and environmental factors being obvious examples). In such situations the BBN must be complemented by other decision making techniques such as those of multi-criteria decision aid (MCDA). In this article we explain the role of BBNs in such decision-making and describe a generic decision-making procedure that uses BBNs and MCDA in a complementary way. The procedure consists of identifying the objective and perspective for the decision problem, as well as the stakeholders. This in turn leads to a set of possible actions, a set of criteria and constraints. We distinguish between, uncertain and certain criteria. The BBN links all the criteria and enables us to calculate a value (within some probability distribution in the case of the uncertain criteria) for each criterion for a given action. This means that we can apply traditional MCDA techniques to combine the values for a given action and then to rank the set of actions. The techniques described are demonstrated by real examples, including a safety assessment example that is being used by a major transportation organisation.

An expert system for the composition of formal Spanish poetry

Abstract: The present paper presents an application that composes formal poetry in Spanish in a semiautomatic interactive fashion. Automatic Spanish Poetry Expert and Rewriting Application (ASPERA) is a forward reasoning rule-based system that obtains from the user basic style parameters and an intended message; applies a knowledge-based pre-processor to select the most appropriate metric structure for the user's wishes; and, by intelligent adaptation of selected examples from a corpus of verses, carries out a prose-to-poetry translation of the given message. In the composition process, ASPERA combines natural language generation and CBR techniques to apply a set of construction heuristics obtained from formal literature on Spanish poetry. If the user validates the poem draft presented by the system, the resulting verses are analysed and incorporated into the system data files.

More than just a pretty face: conversational protocols and the affordances of embodiment

Abstract: Prior research into embodied interface agents has found that users like them and find them engaging. However, results on the effectiveness of these interfaces for task completion have been mixed. In this paper, we argue that embodiment can serve an even stronger function if system designers use actual human conversational protocols in the design of the interface. Communicative behaviors such as salutations and farewells, conversational turn-taking with interruptions, and describing objects using hand gestures are examples of protocols that all native speakers of a language already know how to perform and can thus be leveraged in an intelligent interface. We discuss how these protocols are integrated into Rea, an embodied, multi-modal interface agent who acts as a real-estate salesperson, and we show why embodiment is required for their successful implementation.

An early warning system for loan risk assessment using artificial neural networks

Abstract: The purpose of this paper is to examine the application of artificial neural networks (ANN) to an early warning system for commercial bank loan risk. The aspects of knowledge acquisition, knowledge representation and knowledge-based reasoning process in the early warning system are discussed in the paper. The paper demonstrates how to design a decision tool based on the BP model and provides examples. The results show that this method offers a good tool for early warning.

What Semiotics can and cannot do for HCI

Abstract: Semiotics is 'the mathematics of the humanities' in the sense that it provides an abstract language covering a diversity of special sign-usage (language, pictures, movies, theatre, etc.). In this capacity, Semiotics is helpful for bringing insights from older media to the task of interface design, and for defining the special characteristics of the computer medium. However, Semiotics is not limited to interface design but may also contribute to the proper design of program texts and yield predictions about the interaction between computer systems and their context of use.

Feature space theory — a mathematical foundation for data mining

Abstract: In data mining, an important task in classification and prediction includes feature construction, feature description, feature selection, feature relevance analysis and feature reduction. In this paper, feature space theory is introduced as a mathematical foundation for feature related concepts and techniques in data mining.

Container stowage pre-planning: using search to generate solutions, a case study

Abstract: Container-ships are vessels possessing an internal structure that facilitates the handling of containerised cargo. At each port along the vessel's journey, containers destined for those ports are unloaded and additional containers destined for subsequent ports are loaded. Determining a viable arrangement of containers that facilitates this process, in a cost-effective way, constitutes the deep-sea container-ship stowage problem. This paper outlines a computer system that generates good sub-optimal solutions to the stowage pre-planning problem. This is achieved through an intelligent analysis of the domain allowing the problem to be divided into sub-problems: a generalised placement strategy and a specialised placement procedure. This methodology progressively refines the arrangement of containers within the cargo-space of a container ship until each container is specifically allocated to a stowage location. Good, if not optimal, solutions for the problem are obtained in a reasonable processing time through the use of heuristics incorporated into objective functions for each stage.

Formal concept analysis as a support technique for CBR

Abstract: This paper shows how the use of Galois lattices and formal concept analysis (FCA) can support CBR application designers, in the task of discovering knowledge embedded in the cases. FCA applied on a case library provides an internal sight of the conceptual structure and allows finding patterns, regularities and exceptions among the cases. Moreover, it extracts certain dependence rules between the attributes describing the cases, that will be used to guide the query formulation process.

A new type of recurrent fuzzy neural network for modeling dynamic systems

Abstract: In this paper, a new type of neural network called recurrent fuzzy neural network (RFNN) is proposed to model the fuzzy dynamical systems (FDS). FDS is considered as an order system. The network developed in this paper is based on recurrent neural networks (RNN) to capture the dynamical properties of FDS. The training algorithm is derived based on the tool of order derivative. An example is given to demonstrate the validity of the approach.

A semiotic engineering approach to user interface design

Abstract: Designing software involves good perception, good reasoning, and a talent to express oneself effectively through programming and interactive languages. Semiotic theories can help HCI designers increase their power to perceive, reason and communicate. By presenting some of the results we have reached with semiotic engineering over the last few years, we suggest that the main contributions of semiotic theory in supporting HCI design are: to provide designers with new perceptions on the process and product of HCI design; to bind together all the stages of software development and use, giving them a unique homogeneous treatment; and to pose innovative questions that extend the frontiers of HCI investigations.

The new HCI? navigation of information space

Abstract: When we use the term ‘human–computer interaction’ (HCI), the image that is conjured up is of a person sitting at a visual display unit staring in at the world of ‘information’; the person is very much outside the space of information But when we think of other activities such as going shopping, having a meeting or driving across town, we do not think of the person as outside this space On the contrary, we see the person as inside a space of activities, surrounded by, and interacting with, assorted artefacts and people Navigation of Information Space is an alternative conceptualisation of HCI that sees people as existing inside information spaces Looking at HCI in this way means looking at HCI design as the creation of information spaces This paper explores these ideas in more detail, arguing that Navigation of Information Space is not just a metaphor for HCI; it is a ‘paradigm shift’ The paper illustrates how Semiotics has informed this conception and discusses why such a paradigm shift is needed

Certainty-factor-like structures in Bayesian belief networks

Abstract: The certainty-factor model was one of the most popular models for the representation and manipulation of uncertain knowledge in the early rule-based expert systems of the 1980s. After the model was criticised by researchers in artificial intelligence and statistics as being ad-hoc in nature, researchers and developers have stopped looking at the model. Nowadays, it is often stated that the model is merely interesting from a historical point of view. Its place has been taken over by more expressive formalisms for the representation and manipulation of uncertain knowledge, in particular, by the formalism of Bayesian belief networks. In this paper, it is shown that this view underestimates the importance of the principles underlying the certainty-factor model. In particular, it is shown that certainty-factor-like structures occur frequently in practical Bayesian network models as causal independence assumptions. In fact, the noisy-OR and noisy-AND models, two probabilistic models frequently employed, appear to be reinventions of combination functions previously introduced as part of the certainty-factor model. This insight may lead to a reappraisal of the certainty-factor model.

Enhancing data exploration with a branching history of user operations

Abstract: Backtracking and investigating alternative scenarios are integral parts of exploratory data analysis. Yet today's interfaces cannot represent alternative exploration paths as a branching history, forcing the user to recognize conceptual branch points in a linear history. Further, the interface can only show information from one state at a time, forcing users to rely on memory to compare scenarios. Our system includes a tree-structured visualization for navigating across time and scenarios. The visualization also allows browsing the history and selectively undoing/redoing events within a scenario or across scenarios. It uses the AI formalism of contexts to maintain multiple, possibly mutually inconsistent, knowledge base states. Cross-context formulas can be written for explicit scenario comparison, including visualizations of scenario differences.

Information visualization for intelligent decision support systems

Abstract: To work efficiently with decision support systems (DSS), most users benefit from representation conversion, i.e. translating the specific outcome from the DSS, normally portrayed in a numerical format, into the universal language of the visual. In general, interpretation of data is much more intuitive if the results from the DSS are translated into charts, maps, and other graphical displays because visualization exploits our natural ability to recognize and understand visual patterns. In this paper we discuss the concept of visualization user interface (VUI) for DSS. A proprietary software system known as AniGraftool is introduced as an example of an information visualization application for DSS. In addition, a visualized information retrieval engine based on fuzzy control is proposed.

An agent-based geographical information system

Abstract: In this paper, a prototype geographical information system (GIS) using agent-based technology is introduced. The system known as GXGIS uses agent-based architecture for interactability and scalability. An interface agent is developed to interactively assist the user with the query formation process. The agent has the capability of offering domain knowledge that is associated with a given query. The agent is also able to record a user's troubleshooting experience and show it to the same user or other users as hints. Furthermore, an agent-based architecture with a two-tier mediator model in GIS is developed to satisfy the needs for interoperability among diverse application domains and integration of heterogeneous data sources. As a result, both qualitative and quantitative data can be integrated into GIS.

An object-oriented intelligent design tool to aid the design of manufacturing systems

Abstract: Manufacturing systems design appears to be an excellent area for applying artificial intelligence (AI), object-oriented techniques and computer simulation methods, however, so far there have been limited research on integrating AI, object-oriented techniques and simulation techniques for manufacturing systems design. In this paper, an integrated approach for manufacturing systems design is introduced which consists of the developing and testing of alternative designs using an integrated system SBDSS. SBDSS mainly consists of two subsystems, Object Library Modeler and Simulation Engine with its manager. Using SBDSS, decision makers (DMs) can evaluate alternatives in manufacturing and production, e.g. annual production plan under certain circumstances through scenario simulations. DMs can also test a set of virtual organizations reflecting business process reengineering before a real change is made, as well as testing a virtual manufacturing process for a newly designed product. The paper describes three possible application cases including centralized versus decentralized organizational structure, global goal-oriented planning and highly integrated information systems to illustrate the flexibility of the system.

When actions have consequences: empirically based decision making for intelligent user interfaces

Abstract: One feature of intelligent user interfaces is an ability to make decisions that take into account a variety of factors, some of which may depend on the current situation. This article focuses on one general approach to such decision making: predict the consequences of possible system actions on the basis of prior empirical learning, and evaluate the possible actions, taking into account situation-dependent priorities and the tradeoffs between the consequences. This decision-theoretic approach is illustrated in detail with reference to an example decision problem, for which models for decision making were learned from experimental data. It is shown how influence diagrams and methods of decision-theoretic planning can be applied to arrive at empirically well-founded decisions. This paradigm is then compared with two other paradigms that are often employed in intelligent user interfaces. Finally, various possible ways of learning (or otherwise deriving) suitable decision-theoretic models are discussed.

Presenting through performing: on the use of multiple lifelike characters in knowledge-based presentation systems

Abstract: Trying to imitate the skills of human presenters, some R&D projects have begun to deploy animated characters in presentation tasks. Based either on cartoon drawings, recorded video images of persons, or 3D body models, presentation agents enrich the repertoire of available presentation styles which can be used to effectively communicate information to the user. In this paper, we investigate a new style for presenting information. We introduce the notion of presentation teams which — rather than addressing the user directly — convey information in the style of performances to be observed by him or her. The paper presents an approach to the automated generation of performances which has been tested in two different application scenarios: car sales dialogues and soccer commentary.

Human–computer interaction viewed as pseudo-communication

Abstract: Semiotics is considered fundamental to an understanding of human–computer interaction, and of all computer artifacts. Informatics should therefore be viewed as technical semiotics (or semiotics engineering). In particular, interaction between human and computer is characterized by features of communication, a sort of communication, however, that lacks decisive communicative features. It must be identified as a process of pseudo-communication. Interaction is viewed as the coupling of two autonomous processes: a sign process (carried out by the human user) and a signal process (carried out by the computer). Software appears as a semiotic entity in a duplicate way: calculated and calculating, i.e. both as a result and agent of calculations. This dialectics characterizes the class of signs on the computer medium. Problems of software design (functionality and usability design) are specific problems of the coupling of sign and signal processes.

Smart radio — community based music radio

Abstract: This paper describes the development of a networked music application at Trinity College Dublin. Smart radio is a web-based application, which uses streaming audio technology and collaborative recommendation techniques to allow listeners build, manage and share music programmes. While it is generally acknowledged that music distribution over the web will dramatically change how the music industry operates, there are few prototypes available to demonstrate how this could work in a managed way. The smart radio approach is to have people manage their music resources by putting together personalised music programmes. These programmes can then be swapped using techniques of collaborative recommendation to find similarities between users. The smart radio system currently runs within the computer science intranet with permission from the Irish music rights organisation (IMRO). It is a prototype system for an 'always on' high bandwidth Internet connection such as asymmetric digital subscriber lines (ADSL).

An integrated knowledge-based system for grasslands ecosystems

Abstract: In this study, an integrated knowledge-based system (KBS) for the hilly, sandy grassland ecosystems was developed. The system is a component of the Inner Mongolia Grassland Ecosystem Research Information Systems (IMGERIS). IMGERIS is an important subsystem of Chinese Ecological Research Network (CERN). The system was designed to manage land use, planting species/variety, and optimal coverage of plants. The system integrated a KBS for ecosystem with a model system that includes theoretical models of soil water dynamics. The KBS in the system was coded in Prolog and the model system for soil water dynamics was coded in C. The model system was integrated with the KBS.

Development of a KBS for managing bank loan risk

Abstract: The knowledge-based system (KBS) is introduced as an AI technique applied to bank loan risk management with an emphasis on the study of internal control of bank loans. Descriptions of the design and implementation of KBS for the internal supervising and auditing of bank loans are also included. The structural analysis involved in the system's decision situation is presented and leads to three basic decision components consisting of the auditing of loan approval, the auditing of pre-loan analysis, and the auditing of post-loan supervision. On the basis of these analyses, the structured situation diagram model, the dependency diagram, and the documents needed by the KBS are developed. The system is then implemented using JAVA. Using sample loan data for testing, the KBS prototype is demonstrated to be effective for evaluating the quality of auditing commercial loans.

An object-oriented intelligent CAD system for ceramic kiln

Abstract: This paper describes an integrated intelligent computer-aided design (CAD) system for the ceramic kiln, with an object-oriented database as its core, utilizing object-oriented approach. In addition to the structure and methods of the object-oriented database, the knowledge-based system and the drafting system are discussed. The idea of an integrated intelligent CAD environment is introduced, which can greatly improve design efficiency and quality.

Scheduling railway freight cars

Abstract: It is difficult to allocate the freight cars and to determine how to load the several varieties of goods into them. In order to improve efficiency, a knowledge-based system (KBS) is developed and integrated into the existing railway information system. This paper introduces the development of the system and the framework in which the KBS are integrated into the existing system environments. Under the framework, the combination of logic programming technology and database technology provides us with a practical way to solve the problems in many existing information systems. Meanwhile, developing the KBS as a distributed application benefits both the organization and users. The existing centralized database management system was maintained, and the security and the control of the database were ensured. Users can obtain the performance of a workstation with the transparency of data location in a user-friendly environment. At last, the implementation of the KBS for railway freight loading is discussed.

Extending software through metaphors and metonymies

Abstract: This article is about applications that can be customized or extended through their own user interface. This is achieved by the interface's ability to interpret users' non-literal expressions, namely metaphorical and metonymic ones. Such increased interpretive intelligence depends on static and dynamic models of the domain and application, from which new figurative meanings are abducted automatically or semi-automatically. The system performs controlled modifications on the underlying models, based on inferences it draws about users' intentions as they produce figurative utterances.

An architecture for developing attentive information systems

Abstract: Attentive systems attend to what users do so that they can attend to what users need. Such systems track user behavior, model user interests, and anticipate user desires and actions. Because the general class of attentive systems is broad — ranging from human butlers to web sites that profile users — we have focused specifically on attentive information systems, which observe user actions with information resources, model user information states, and suggest information that might be helpful to users. In particular, we describe Simple User Interest Tracker (Suitor), an architecture for developing attentive information systems that track computer users through multiple channels — eye gaze, web browsing, application use, to determine interests and to try to satisfy information needs. By observing behavior and modeling users, Suitor can be used to find and display potentially relevant information that is both timely and non-disruptive to the user's ongoing activities.

Competence guided incremental footprint-based retrieval

Abstract: Case-based reasoning (CBR) systems solve new problems by retrieving and adapting problem solving experiences stored as cases in a case-base. Success depends largely on the performance of the case retrieval algorithm used. Smyth and McKenna [Lecture Notes in Artificial Intelligence LNAI 1650 (1999) 343–357] have described a novel retrieval technique, called footprint-based retrieval (FBR), which is guided by a model of case competence. FBR as it stands benefits from superior efficiency characteristics and achieves near-optimal competence and quality characteristics. In this paper, we describe a simple but important extension to FBR. Empirically we show that this new algorithm can deliver optimal retrieval performance while at the same time retaining the efficiency benefits of the original FBR method.