Possibilistic Similarity Measures

Dual coding of knowledge in the human brain.

Math anxiety is a negative emotional response arising in mathematical contexts that hinders mathematical performance. It not only impacts students' math achievements but also exerts a profound influence on their mental health and life choices. To clarify the correlation characteristics between different representation forms of high school mathematics knowledge and math anxiety, and to provide empirical evidence for alleviating students' math anxiety through optimizing knowledge representation strategies in teaching, this study selected 300 high school students from an ordinary high school in a city in China, covering three grades (Grade 10 to Grade 12), including 116 male students and 174 female students. Based on the theory of mathematical multiple representations, this study classified the representation forms of high school mathematics knowledge into four types: symbolic representation, verbal representation, graphical representation, and situational representation. Data were collected through a questionnaire survey, and empirical research was conducted using methods such as descriptive statistical analysis, partial correlation analysis, and multiple regression analysis. The results show that all different knowledge representation forms are positively correlated with math anxiety, among which graphical representation and verbal representation have a strongly positive correlation with math anxiety, indicating that these two forms are more likely to induce students' math anxiety. Meanwhile, the overall level of math anxiety can be effectively predicted based on students' anxiety levels under different knowledge representation forms, with a good prediction effect. This study can provide precise intervention strategies for high school mathematics teaching.

We present the Angerona framework for the implementation of knowledge-based agents with a strong focus on flexibility, extensibility, and compatibility with diverse knowledge representation formalisms. As the basis for this framework we propose and formalize a general concept of compound agents in which we consider agents to consist of hierarchies of interacting epistemic and functional components. Each epistemic component is instantiated by a knowledge representation formalism. Different knowledge representation formalisms can be used within one agent and different agents in the same system can be based on different agent architectures and can use different knowledge representation formalisms. Partially instantiations define sub-frameworks for, e. g., the development of BDI agents and variants thereof. The Angerona framework realizes this concept by means of a flexible JAVA plug-in architecture for the epistemic and the functional components of an agent. The epistemic plug-ins are based on the Tweety library for knowledge representation, which provides various ready-for-use implementations and knowledge representation formalisms and a framework for the implementation of additional ones. Angerona already contains several partial and complete instantiations that implement several approaches. Angerona also features an environment plug-in for communicating agents and a flexible GUI to monitor the multiagent system and the inner workings of the agents, particularly the inspection of the dynamics of their epistemic states. Angerona and Tweety are ready to use, well documented, and open source.KeywordsKnowledge RepresentationMultiagent SystemEpistemic StateBelief BaseBelief ChangeThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Decision trees are among the most popular classification algorithms due to their knowledge representation in form of decision rules which are easy for interpretation and analysis. Nonetheless, a majority of decision trees training algorithms base on greedy top-down induction strategy which has the tendency to develop too complex tree structures. Therefore, they are not able to effectively generalise knowledge gathered in learning set. In this paper we propose EVO-Tree hybrid algorithm for decision tree induction. EVO-Tree utilizes evolutionary algorithm based training procedure which processes population of possible tree structures decoded in the form of tree-like chromosomes. Training process aims at minimizing objective functions with two components: misclassification rate and tree size. We test the predictive performance of EVO-Tree using several public UCI data sets, and we compare the results with various state-of-the-art classification algorithms.

THINGS-data reflects three large-scale neuroimaging and behavioral datasets of object processing in humans, comprising densely sampled functional MRI and magnetoencephalographic recordings, as well as 4.70 million similarity judgments in response to thousands of photographic images for up to 1854 objects.

The paper is devoted to resolving a topical issue of automatized tests production in poor-formalized knowledge domains. An approach was proposed to carry out an automatized generation of multiple-choice tests on the basis of an ontological model of a domain based on prototypes of human cognitive structures. The universal ontological three-level model was developed with the core element called "cognitive prototype" as a data structure for knowledge representation in form of such known cognitive structures as concept, frame, scheme, scenario, etc. The formal description of test patterns was devised based on cognitive prototypes as the first step towards their typification, developing of algorithms of their automatized generation and programming of subsystem of intelligent educational systems. Prospects of such approach have been analyzed for designing unified techniques of tests production with reduction of laboriousness of the test preparation process.

Interval-valued intuitionistic fuzzy numbers (IVIFNs) are better in modelling real-life problems more naturally, and it can apply to many fields such as Pattern Recognition, Decision Making, Cluster Analysis, Medical Diagnosis, Image Processing, etc. Especially similarity measures defined on the class of IVIFNs does play a significant role in the field of Pattern Recognition and Decision Making. Many authors from all over the world were trying to define a standard method (Similarity measure) that can be suitable for most of the problems. Unfortunately, each of the similarity measures has certain drawbacks as well as some advantages among different similarity measures available in the literature. Most of the similarity measures defined in the particular class (subset) of IVIFNs. These issues open up a pathway for further/ future research. In this study, we aim at introducing a new similarity measure defined in another class of an IVIFNs that can cover more IVIFNs in it. In this paper, first, we define a new similarity measure on the class of interval-valued intuitionistic fuzzy numbers (IVIFNs) based on the non-hesitance score function defined on the class of IVIFNs. Second, we discuss the drawback of various existing similarity measures and compare them with the proposed similarity measures using different cases. Third, the efficacy of the proposed similarity measure to familiar existing methods is studied using illustrative examples. Finally, the applicability of the proposed method in solving pattern recognition problem is depicted.

Intensional dynamic programming. A Rosetta stone for structured dynamic programming

Knowledge representation form in mechanical engineering

Chapter 1 Knowledge Representation and Classical Logic

Fuzzy Petri nets for knowledge representation and reasoning: A literature review

The current knowledge about biochemical networks is largely incomplete. Thus biologists constantly need to revise or extend existing knowledge. These revision or extension are first formulated as theoretical hypotheses, then verified experimentally. Recently, biological data have been produced in great volumes and in diverse formats. It is a major challenge for biologists to process these data to reason about hypotheses. Many computer-aided systems have been developed to assist biologists in undertaking this challenge. The majority of the systems help in finding “pattern” in data and leave the reasoning to biologists. Few systems have tried to automate the reasoning process of hypothesis formation. These systems generate hypotheses from a knowledge base and given observations. A main drawback of these knowledge-based systems is the knowledge representation formalism they use. These formalisms are mostly monotonic and are now known to be not quite suitable for knowledge representation, especially in dealing with incomplete knowledge, which is often the case with respect to biochemical networks. We present a knowledge based framework for the general problem of hypothesis formation. The framework has been implemented by extending BioSigNet-RR. BioSigNet-RR is a knowledge based system that supports elaboration tolerant representation and non-monotonic reasoning. The main features of the extended system include: (1) seamless integration of hypothesis formation with knowledge representation and reasoning; (2) use of various resources of biological data as well as human expertise to intelligently generate hypotheses. The extended system can be considered as a prototype of an intelligent research assistant of molecular biologists. The system is available at http://www.biosignet.org.

Similarity and Difference Judgments Under Perceptual and Non-Perceptual Conditions Uri Hasson (uhasson@princeton.edu) Department of Psychology, Princeton University Princeton, NJ 08544, USA Vladimir Sloutsky (Sloutsky.1@osu.edu) Center for Cognitive Science & School of Teaching and Learning, The Ohio State University 21 Page Hall, 1810 College Road, Columbus, OH 43210, USA Abstract It has recently been suggested that knowledge is represented in the form of perceptual symbol systems (Barsalou, 1999). According to this view, perceptual states may be used to support higher cognitive processes without being transduced into a representational language. Since the ability to recognize difference and similarity is fundamental for cognition, we examined to what extent it might be based on perceptual information. In three experiments, participants made judgments of similarity and difference for simple items under three presentation conditions: Words Only, Words and Pictures, and Pictures Only. Reaction times for judgments in the Words-Only condition were consistently slower than in the other presentation conditions. However, judgments of perceived similarity and perceived difference did not markedly differ between presentation conditions. The results suggest that participants recruited perceptual information when evaluating similarity in the Word-Only condition. Additionally, presentation condition had an effect on the relation between the similarity and difference scales: for a given degree of similarity, more extreme difference judgments were found under those conditions where words were displayed. We offer an explanation for this effect, and present a further research program. Introduction Similarity, or psychological resemblance of entities, is a fundamental aspect of cognition. Similarity plays a critical role in perception, memory, learning and transfer, categorization, analogical reasoning, problem solving, and language comprehension. It has also been suggested that recognizing differences between entities is fundamental to cognition; for instance, in discriminating category members from non-members. An extensive body of research then has outlined the processes for which similarity and difference are important. However, much less is known about that nature of the information recruited for evaluating similarity and difference themselves. Some theories of similarity hypothesize a process whose inputs are representations in the form of feature lists with most of the properties represented as unitary attributes (Nosofsky, 1986; Tversky, 1977). According to others, conceptual knowledge plays such an important role in similarity, that similarity is taken to be akin to analogy; operating on an interconnected system of relations and their arguments (Gentner & Markman, 1997). The crux of this view is that processes of similarity, as well as difference (Markman, 1996), operate by aligning the structures of the compared entities. The structures themselves are described as a system of relational predicates and their attributes. The research program of structural alignment (see Gentner & Markman, 1997) convincingly demonstrated that the use of structural knowledge is an integral part of making similarity judgments. It is an open question, however, whether the only kinds of information used for these judgments are in the form of such a-modal representations as relations or feature lists. A recent proposal (Barsalou, 1999), suggests that perceptual states - modal and analog forms of representation - are also used to support higher cognitive processes. Perceptual states are taken to maintain a part of the perceptual nature of their referents, without being transduced into a representational language. If this is the case, judgments of similarity and/or difference might make use of perceptual information, in addition to other forms of knowledge. We examined this issue by evaluating the effects of various types of presentation conditions on similarity and difference judgments, which were made for a variety of natural stimuli. We presented participants with pairs of items for similarity and difference judgments under three presentation conditions which were manipulated between groups: pairs were presented as Words-Only (WO), as words accompanied by pictures (WP) or as pictures only (PO). The instructions given to all three groups were the same, and did not make any reference to speed of response. Most critically, the instructions given to participants in the Word-Only condition did not mention envisioning the objects depicted by the words. If participants in this

The measurement of uncertainty is an important topic for the theories dealing with uncertainty. The definition of similarity measure between two IFSs is one of the most interesting topics in IFSs theory. A similarity measure is defined to compare the information carried by IFSs. Many similarity measures have been proposed. A few of them come from the well-known distance measures. In this work, a new similarity measure between IFSs was proposed by the consideration of the information carried by the membership degree, the non-membership degree, and hesitancy degree in intuitionistic fuzzy sets (IFSs). To demonstrate the efficiency of the proposed similarity measure, various similarity measures between IFSs were compared with the proposed similarity measure between IFSs by numerical examples. The compared results demonstrated that the new similarity measure is reasonable and has stronger discrimination among them. Finally, the similarity measure was applied to pattern recognition and medical diagnosis. Two illustrative examples were provided to show the effectiveness of the pattern recognition and medical diagnosis.

To explore the strengths and pitfalls of mapping structured EPR (electronic patient record) terms (OpenSDE) to SNOMED codes. The OpenSDE model was developed for cardiovascular diseases in the context of the I4C project. We employed 35 patient records as references to adjust the model. We then performed automated and manual matches following the design of matching terms in the thesaurus of the resulting OpenSDE domain model to SNOMED concepts. Subsequently, we assessed what number of OpenSDE terms within the domain model can be matched to SNOMED concepts. The OpenSDE domain tree contains 3230 nodes, involving 689 unique terms (terms can be associated with more than one node in different parts of a domain model tree). After final manual work for the 689 tree terms, 616 resulted in a good match, 31 in a partial match, and 42 in no match. Of the good matches, 23 produced multiple matches. The matches were used to represent the mapping of each node in the domain tree by concatenation of the matching terms. Mapping predefined terms in OpenSDE domain models to SNOMED Clinical Terms (CT) concepts eliminates laborious mapping for each individual patient record. The assignment of SNOMED codes to OpenSDE tree nodes facilitates exchange, aggregation, and research involving patient data. The mapping will serve the construction of queries at higher semantic levels than explicitly modeled in an OpenSDE domain model. However, the usefulness of the mapping result depends on the completeness of the mapping to SNOMED CT, for which there is no gold standard.