Showing papers by "Turku Centre for Computer Science published in 2004"
TL;DR: This paper combines data mining methods for analyzing quantitative and qualitative data from financial reports, in order to see if the textual part of the report contains some indication about future financial performance.
Abstract: There is a vast amount of financial information on companies’ financial performance available to investors today. While automatic analysis of financial figures is common, it has been difficult to automatically extract meaning from the textual part of financial reports. The textual part of an annual report contains richer information than the financial ratios. In this paper, we combine data mining methods for analyzing quantitative and qualitative data from financial reports, in order to see if the textual part of the report contains some indication about future financial performance. The quantitative analysis has been performed using selforganizing maps, and the qualitative analysis using prototype-matching text clustering. The analysis is performed on the quarterly reports of three leading companies in the telecommunications sector.
159 citations
TL;DR: This work compares the number of states between minimal deterministic finite automata accepting a regular language and its reversal and presents several classes of languages where this maximal blow-up is actually achieved.
117 citations
TL;DR: A measure of interactivity between marginal distributions of a joint possibility distribution C is introduced as the expected value of the interactivity relation between the γ-level sets of its marginal distributions.
114 citations
TL;DR: This paper introduces the closely related notion of a subword history and obtains a sequence of general results: elimination of products, decidability of equivalence, and normal form, and investigates overall-methods for proving the validity of such results.
90 citations
TL;DR: This article gives an overview of artificial neural network studies conducted in the auditing field and argues that these auditing ANN‐applications could serve the analytical review (AR) process.
Abstract: This article gives an overview of artificial neural network (ANN) studies conducted in the auditing field. The review pays attention to application domains, data and sample sets, ANN‐architectures and learning parameters. The article argues that these auditing ANN‐applications could serve the analytical review (AR) process. The summary of the findings pays attention to whether authors state that ANNs have potential to improve analytical review (AR) procedures. Furthermore, the article evaluates which are the most influential contributions and which are open ends in the field. The article makes some practical suggestions to motivate academics and practitioners to collaborate in further exploration of the potential of ANNs.
78 citations
TL;DR: The paper investigates inequalities between the numbers of different (scattered) subword occurrences and gives various characterizations for Parikh matrices in the case where the matrix determines the word uniquely.
Abstract: The paper investigates inequalities between the numbers of different (scattered) subword occurrences. The Parikh matrix recently introduced is an efficient tool. We give various characterizations for Parikh matrices. Of special interest is the case where the matrix determines the word uniquely. We investigate such matrix unambiguous words. The considerations are extended to concern languages. Several open problems and problem areas are indicated.
56 citations
01 Sep 2004
TL;DR: This paper proposes an efficient image‐based approach for computing soft shadows based on shadow mapping and uses pixel‐based visibility computations for rendering penumbra regions directly into the screen‐space using programmable graphics hardware.
Abstract: Most former soft shadow algorithms have either suffered from restricted self-shadowing capabilities, been too slow for interactive applications, or could only be used with a limited types of geometry. In this paper, we propose an efficient image-based approach for computing soft shadows. Our method is based on shadow mapping and provides the associated benefits. We use pixel-based visibility computations for rendering penumbra regions directly into the screen-space. This is accomplished by using a modified flood-fill algorithm which enables us to implement the algorithm using programmable graphics hardware. Even though the resulting images are most often high quality, we do not claim that the proposed method is physically correct. The computation time and memory requirements for soft shadows depend on image resolution and the number of lights, not geometric scene complexity.
Categories and Subject Descriptors (according to ACM CCS): I.3.3 [Computer Graphics]: Picture/Image Generation I.3.7 [Three-Dimensional Graphics and Realism]: Color, shading, shadowing, and texture
44 citations
Journal Article•
TL;DR: A condition introduced in this paper, called γ-property, turns out to be a powerful tool for such unambiguous matrices, especially the case where the matrix determines the word uniquely.
Abstract: We investigate the number of (scattered) subword occurrences and Parikh matrices, especially the case where the matrix determines the word uniquely. A condition introduced in this paper, called γ-property, turns out to be a powerful tool for such unambiguous matrices. Interconnections with the general theory of subword histories are also pointed out.
41 citations
TL;DR: It is proved that the set of rough sets determined by a symmetric and transitive binary relation forms a complete Stone lattice and some possible canonical completions are presented.
Abstract: We study the ordered set of rough sets determined by relations which are not necessarily reflexive, symmetric, or transitive. We show that for tolerances and transitive binary relations the set of rough sets is not necessarily even a semilattice. We also prove that the set of rough sets determined by a symmetric and transitive binary relation forms a complete Stone lattice. Furthermore, for the ordered sets of rough sets that are not necessarily lattices we present some possible canonical completions.
31 citations
TL;DR: The Banerjee definition of a fuzzy choice function (=fuzzy consumer) is modified and some fuzzy versions of the axioms of revealed preference and congruence are studied, establishing some connections between WAFRP, SAFRP, WFCA, SFCA extending a significant part of Sen theorem.
Abstract: The revealed preference is a central subject in classical consumer theory. Authors like Samuelson, Arrow, Richter, Sen, Uzawa and others have proposed an axiomatic setting of revealed preference theory.
29 citations
TL;DR: A survey and a unified presentation of the defect theorem, its generalizations and recent aspects of interest, which are related to equations of words, and in this way to the compactness theorem for systems of word equations.
24 May 2004
TL;DR: A model-based approach that combines the data-flow and object-oriented computing paradigms to model embedded systems and shows that by using models it is able to implement automated transformations between different views of the system under design.
Abstract: This paper presents a model-based approach that combines the data-flow and object-oriented computing paradigms to model embedded systems. The rationale behind the approach is that both views are important for modelling purposes in embedded systems environments, and thus a combined and integrated usage is not only useful, but also fundamental for developing complex systems. We also show that by using models we are able to implement automated transformations between different views of the system under design. We exemplify the approach with an IPv6 router case study.
TL;DR: Data and text mining methods are combined for analysing quantitative and qualitative data from fnancial reports, in order to see if the textual part of the report contains some indications about future fnancial performance.
Abstract: There is a vast amount of fnancial information on companies' fnancial performance available to investors in electronic form today. While automatic analysis of fnancial fgures is common, it has been diffcult to extract meaning from the textual parts of fnancial reports automatically. The textual part of an annual report contains richer information than the fnancial ratios. In this paper, we combine data and text mining methods for analysing quantitative and qualitative data from fnancial reports, in order to see if the textual part of the report contains some indications about future fnancial performance. The quantitative analysis has been performed using self-organizing maps, and the qualitative analysis using prototype-matching text clustering. The analysis is performed on the quarterly reports of three leading companies in the telecommunications sector. Copyright © 2004 John Wiley & Sons, Ltd.
01 Apr 2004
TL;DR: It is shown that finite, unbordered factors of Sturmian words are Lyndon words and that, if wu is a minimal Duval extension, then u is a factor of w.
Abstract: A word v=wu is a (nontrivial) Duval extension of the unbordered word w, if (u is not a prefix of v and) w is an unbordered factor of v of maximum length. After a short survey of the research topic related to Duval extensions, we show that, if wu is a minimal Duval extension, then u is a factor of w. We also show that finite, unbordered factors of Sturmian words are Lyndon words.
Journal Article•
TL;DR: The consistency conditions Fα, Fβ, Fδ as fuzzy forms of Sen's properties alpha, beta and delta are introduced and it is shown that a fuzzy consumer is normal if and only if conditions F α2, Fγ2 hold.
Abstract: This paper is concerned with consistency properties for fuzzy consumers. We introduce the consistency conditions Fα, Fβ, Fδ as fuzzy forms of Sen's properties alpha, beta and delta. Other consistency conditions are also studied: Fα2, Fβ(+), Fγ2 and FPI. The main results are:
(1) A fuzzy consumer verifies Fα, Fβ if and only if the congruence axiom WFCA holds;
(2) If h is a normal fuzzy consumer then Fδ holds if and only if the associated fuzzy preference relation R is quasi-transitive;
(3) A fuzzy consumer is normal if and only if conditions Fα2, Fγ2 hold.
26 May 2004
TL;DR: This work motivates a study for the convergence of the two overlay network technologies and sketches an evolving architecture towards integrating the two technologies in building overlay network applications.
Abstract: Peer-to-peer (P2P) networks and mobile ad hoc networks (MANET) share some key characteristics: self-organization and decentralization, and both need to solve the same fundamental problem: connectivity. We motivate a study for the convergence of the two overlay network technologies and sketch an evolving architecture towards integrating the two technologies in building overlay network applications.
TL;DR: A quadratic upper bound for the neighborhood size of the inverse automaton of some types of one-dimensional reversible cellular automata is given, and it is shown that this bound can be lowered in some particular cases.
Journal Article•
TL;DR: A HOL formalisation for analysing expected time bounds for probabilistic programs based on the quantitative program logic of Morgan et al.
Abstract: We provide a HOL formalisation for analysing expected time bounds for probabilistic programs Our formalisation is based on the quantitative program logic of Morgan et al [21] and McIver's extension of it [17] to include performance-style operators In addition we provide some novel results based on probabilistic data refinement which we use to improve the utility of the basic method
Journal Article•
TL;DR: The new revealed preference axioms WAFRP, SAFRP, HAFRP are introduced and two main theorems are proved: the axiomers WFCA and WAF RP$^o$ are equivalent; and theAxioms SFC a and HAF RP are equivalent.
Abstract: In a previous paper connections between the congruence axioms WFCA, SFCA and the revealed preference axioms WAFRP, SAFRP for fuzzy choice functions whose domain contains the characteristic functions of pairs and of triples of alternatives have been studied The first objective of this paper is to establish such connections for the case of arbitrary fuzzy choice functions We introduce the new revealed preference axioms WAFRP$^o$, SAFRP$^o4, HAFRPand we prove two main theorems: 1 The axioms WFCA and WAFRP$^o$ are equivalent;2 The axioms SFCA and HAFRP are equivalent Our second objective is to define G-rational, M-rational, G-normal and M-normal fuzzy choice functions and to investigate some of their properties The notions and the results are formulated in terms of the residuum of the Godel t-norm and the proofs use the residuated lattice structure of the interval [0,1]
TL;DR: It is proved that, for each word w of length n, the sequence w,β(w),β2(w,... terminates either in bn or in the cycle of conjugates of the word abkabk+1 for n = 2k + 3.
TL;DR: This paper compares the two models for gene assembly in ciliates, formalizing both in terms of pointer reductions, and discusses invariants and universality results for both models.
Abstract: Two models for gene assembly in ciliates have been proposed and investigated in the last few years. The DNA manipulations postulated in the two models are very different: one model is intramolecular – a single DNA molecule is involved here, folding on itself according to various patterns, while the other is intermolecular – two DNA molecules may be involved here, hybridizing with each other. Consequently, the assembly strategies predicted by the two models are completely different. Interestingly however, the final result of the assembly (including the assembled gene) is always the same. We compare in this paper the two models for gene assembly, formalizing both in terms of pointer reductions. We also discuss invariants and universality results for both models.
TL;DR: A morphological model is developed which generates spelling and inflection variants of the interaction words of protein-protein interaction sentences in biomedical literature in order to broaden the set of potential interaction words.
Abstract: In this paper, we introduce a way to apply rough set data analysis to the problem of extracting protein-protein interaction sentences in biomedical literature Our approach builds on decision rules of protein names, interaction words, and their mutual positions in sentences In order to broaden the set of potential interaction words, we develop a morphological model which generates spelling and inflection variants of the interaction words We evaluate the performance of the proposed method on a hand-tagged dataset of 1894 sentences and show a precision-recall break-even performance of 79,8% by using leave-one-out crossvalidation
TL;DR: This paper describes how chlorophyll a fluorescence images taken in blue light through a red filter, can be used to classify apples and proposes an apple classification method that employs a hierarchy of two neural networks.
Abstract: Classification of harvested apples when predicting their storage potential is an important task. This paper describes how chlorophyll a fluorescence images taken in blue light through a red filter, can be used to classify apples. In such an image, fluorescence appears as a relatively homogenous area broken by a number of small nonfluorescing spots, corresponding to normal corky tissue patches, lenticells, and to damaged areas that lower the quality of the apple. The damaged regions appear more longish, curved or boat-shaped compared to the roundish, regular lenticells. We propose an apple classification method that employs a hierarchy of two neural networks. The first network classifies each spot according to geometrical criteria and the second network uses this information together with global attributes to classify the apple. The system reached 95% accuracy using a test material classified by an expert for "bad" and "good" apples.
TL;DR: In this article, a job grouping problem with several feeder types (JGP-T) is considered, where instead of one linear feeder, each component is associated with a given feeder type which restricts its placement.
Abstract: A variant of the classical job grouping problem (JGP) in printed circuit board (PCB) assembly is considered. Studies on JGPs have assumed a single feeder from which the components are retrieved and then placed on the PCB. Recent advances in technology have made it possible to use several different kinds (types) of feeders at the same time. In a JGP, the aim is to group the PCBs so that the cardinality of the grouping is minimal and each group can be processed without rearranging the contents of the feeder. In the job grouping problem with several feeder types (JGP-T) the goal is the same but instead of one linear feeder we have several feeders and each component is associated with a given feeder type which restricts its placement. We give a mathematical formulation for the JGP-T and show that it is hard to solve to optimality for problems of practical size. The connections of JGP-T to known problems are discussed. We also propose several efficient heuristics and compare their results against optimal solutions.
15 Sep 2004
TL;DR: A framework for mobile applications aimed at supporting salespersons’ tasks for greater performance when they are operating within a highly mobile work environment is suggested.
Abstract: The paper suggests a framework for mobile applications aimed at supporting salespersons’ tasks for greater performance when they are operating within a highly mobile work environment. To do so the paper starts by providing a review of mobile technologies characteristics in terms of mobile devices, connectivity and mobile applications. After deriving a set of propositions, the paper offers some concluding remarks and suggests areas for future research
TL;DR: A heuristic, multivariate, top‐down partitioning algorithm that divides attribute ranges into such intervals that the discovered frequent sets are also dense, and thus probably more interesting to the user.
Abstract: The article suggests a partitioning algorithm for quantitative attributes to support the discovery of frequent fuzzy patterns among transactions containing such attributes. More precisely, we present a heuristic, multivariate, top-down partitioning algorithm that divides attribute ranges into such intervals that the discovered frequent sets are also dense, and thus probably more interesting to the user. Our approach is fuzzy, so that the derived intervals have fuzzy bounds, and thereby also the derived frequent sets are fuzzy. The crisp (nonfuzzy) case is obtained as a special case. We evaluate the goodness of the partitioning method by measuring the average and absolute information amounts of the obtained fuzzy frequent sets. For the mining task, any fuzzy frequent item set mining method can be used. Experiments show that the algorithm is able to do multidimensional partitioning in a balanced way, and the “interestingness” of the obtained frequent sets is quite high, especially for correlated attributes. © 2004 Wiley Periodicals, Inc. Int J Int Syst 19: 1111–1126, 2004.
01 Jan 2004
TL;DR: In this paper, the Rialto intermediate language is used to capture the semantics of UML behavioral diagrams, which can be used to uniformly describe the behavior of a combination of several diagrams and as a bridge from UML models to animation and production code.
Abstract: In this article we show how to use the Rialto intermediate language, to capture the semantics of UML behavioral diagrams. The Rialto language has a formal semantics given as structural operational rules and it supports semantic variations. It can be used to uniformly describe thc behavior of a combination of several diagrams and as a bridge from UML models to animation and production code.
TL;DR: A more general case by taking unary X-algebras instead of automata is considered and complete lattice isomorphisms are established between the lattices of σ-varieties of X-numbers, κ- varieties of semigroups and weakly invariant congruences on the free semigroup X+, where κ is the cardinality of X.
TL;DR: A novel method for optimizing photomultiplier (PMT) pulse discrimination levels in single- and multi-photon counting is demonstrated and it is applicable to automatic calibration of photon counting devices on production line.
Abstract: In applications where random multi-photon events must be distinguishable from the background, detection of the signals must be based on either analog current measurement or photon counting and multi-level discrimination of single and multi-photon events. In this paper a novel method for optimizing photomultiplier (PMT) pulse discrimination levels in single- and multi-photon counting is demonstrated. This calibration method is based on detection of photon events in coincidence to short laser pulses. The procedure takes advantage of Poisson statistics of single- and mult-iphoton signals and it is applicable to automatic calibration of photon counting devices on production line. Results obtained with a channel photomultiplier (CPM) are shown. By use of three parallel discriminators and setting the discriminator levels according to the described method resulted in a linear response over wide range of random single- and multi-photon signals.
23 Jun 2004
TL;DR: The method avoids duplication of code blocks when compiling SDF graphs regardless of whether a single appearance schedule can be found for the graph or not and also means that the method can use schedules that require minimal buffer memory but are not single appearance schedules.
Abstract: We present a novel approach for efficient code synthesis from synchronous dataflow specifications. The method avoids duplication of code blocks when compiling SDF graphs regardless of whether a single appearance schedule can be found for the graph or not. This also means that we can use schedules that require minimal buffer memory but are not single appearance schedules. The method has been developed within the compiler for the Rialto language, which we have developed for use as an intermediate language for code synthesis from heterogeneous models of computation. The optimization technique presented in the paper can, however, very well also be used without the Rialto language.