Search

Abstract Accurate semantic representation models are essential in text mining applications. For a successful application of the text mining process, the text representation adopted must keep the interesting patterns to be discovered. Although competitive results for automatic text classification may be achieved with traditional bag of words, such representation model cannot provide satisfactory classification performances on hard settings where richer text representations are required. In this paper, we present an approach to represent document collections based on embedded representations of words and word senses. We bring together the power of word sense disambiguation and the semantic richness of word- and word-sense embedded vectors to construct embedded representations of document collections. Our approach results in semantically enhanced and low-dimensional representations. We overcome the lack of interpretability of embedded vectors, which is a drawback of this kind of representation, with the use of word sense embedded vectors. Moreover, the experimental evaluation indicates that the use of the proposed representations provides stable classifiers with strong quantitative results, especially in semantically-complex classification scenarios. [ABSTRACT FROM AUTHOR]

Abstract Extensible Business Reporting Language (XBRL) is an XML-based language developed for enhancing interoperability among the entities involved in process of business reporting. Although this language is adopted by various regulators all around the world and has contributed greatly to semantic interoperability in this field, the variations between taxonomies and also between elements of instance documents, still cause many inconsistencies between elements. Although some existing approaches suppose the conversion of XBRL to ontologies and then resolve the inconsistencies by applying some mapping techniques, it does not seem practical because of low precision and incompleteness of these conversions. In this paper, a novel approach is proposed which utilizes Ant Colony Optimization (ACO) in order to detect best semantic mappings between inconsistent concepts of two XBRL documents. This approach analyzes the possible mappings with respect to various factors like concept names, all label texts, presentation and calculation hierarchies and so on. This makes the approach capable of finding mappings, which were not easily discoverable otherwise. The proposed approach is implemented and applied to actual XBRL reports. The results are measured with aid of well-known criteria (precision, recall and F-measure) and are compared with the well-known Hungarian algorithm and illustrate the better performance in accordance with these three criteria. [ABSTRACT FROM AUTHOR]

As computer science and information technology are making broad and deep impacts on our daily lives, more and more papers are being submitted to computer science journals and conferences. To help authors decide where they should submit their manuscripts, we present the Content-based Journals & Conferences Recommender System on computer science, as well as its web service at http://www.keaml.cn/prs/ . This system recommends suitable journals or conferences with a priority order based on the abstract of a manuscript. To follow the fast development of computer science and technology, a web crawler is employed to continuously update the training set and the learning model. To achieve interactive online response, we propose an efficient hybrid model based on chi-square feature selection and softmax regression. Our test results show that, the system can achieve an accuracy of 61.37% and suggest the best journals or conferences in about 5 s on average. [ABSTRACT FROM AUTHOR]

This paper presents algorithmic and empirical contributions demonstrating that the convergence characteristics of a co-evolutionary approach to tackle Multi-Objective Games (MOGs) with postponed preference articulation can often be hampered due to the possible emergence of the so-called Red Queen effect. Accordingly, it is hypothesized that the convergence characteristics can be significantly improved through the incorporation of memetics (local solution refinements as a form of lifelong learning), as a promising means of mitigating (or at least suppressing) the Red Queen phenomenon by providing a guiding hand to the purely genetic mechanisms of co-evolution. Our practical motivation is to address MOGs characterized by computationally expensive evaluations, wherein there is a natural need to reduce the total number of true evaluations consumed in achieving good quality solutions. To this end, we propose novel enhancements to co-evolutionary approaches for tackling MOGs, such that memetic local refinements can be efficiently applied on evolved candidate strategies by searching on computationally cheap surrogate payoff landscapes (that preserve postponed preference conditions). The efficacy of the proposal is demonstrated on a suite of test MOGs that have been designed. [ABSTRACT FROM AUTHOR]

Most of the existing similarity metrics in heterogeneous information networks depend on the pre-specified meta-path or meta-structure. This dependency may cause them to be sensitive to different meta-paths or meta-structures. In this paper, we propose a stratified meta-structure-based similarity measure named SMSS in heterogeneous information networks. The stratified meta-structure can be constructed automatically and capture rich semantics.Then, we define the commuting matrix of the stratified meta-structure by virtue of the commuting matrices of meta-paths and meta-structures. As a result, the SMSS is defined by virtue of this commuting matrix. Experimental evaluations show that the existing metrics are sensitive to different meta-paths or meta-structures and that the proposed SMSS outperforms the state-of-the-art metrics in terms of ranking and clustering. [ABSTRACT FROM AUTHOR]

This paper presents a model for semantic network representation. According to the model, the main purpose of a semantic network is to help to correctly and quickly understand the features and structure of the solutions, and then make reliable predictions about their outcomes and rewards. These objectives are achieved under constrains like limited resources (e.g., memory size and access), time limitations, and variations of solution characteristics. The paper summarizes the elements of the model and presents a set of properties that describe the effectiveness of semantic networks in creative problem solving, including design. Six examples (based on three case studies from the related literature and three experimental studies conducted by the authors) discuss the model elements, characteristics, and properties. [ABSTRACT FROM AUTHOR]

The past half-century has witnessed fast development in the field of fuzzy theory (FT), however, few researches have focused on mapping the development of this area in China. Based on the samples of 12,936 publications authored by Chinese scholars on FT researches during the past 30 years, this paper intends to explore the patterns and dynamics by analyzing the geographic distribution of publications, international collaboration, research hot spot, subject categories and journals, and publication contributors. The results indicate that the scientific publications are highly unbalanced at regional levels in China, and the USA is China's most important partner in FT cooperative researches. Collaborations are not indispensable for high-quality research outputs in FT area. The existing researches in the field of FT from Chinese scholars focus primarily on Computer Science and Engineering. The emerging trends of FT researches from Chinese scholars have shifted away from basic FT researches to the applications, such as the areas of decision making, optimization, modeling and design. [ABSTRACT FROM AUTHOR]

This paper comments on the published work dealing with "DSKmeans: A new kmeans-type approach to discriminative subspace clustering" [Knowledge-Based Systems, Vol. 70, pp. 293–300, 2014] proposed by X. Huang et al. Their clustering approach is based on a new mathematical model with two groups of variables: cluster centers and clusters memberships. They proposed an iterative algorithm to obtain the solution of this model. In each iteration, they fixed clusters memberships and claimed that the optimal cluster centers can be obtained by setting the derivative of the objective function of the model to zero. In this paper, we show that their proposed method cannot obtain the optimal solution of cluster centers for any given clusters memberships and some values of the model parameter. [ABSTRACT FROM AUTHOR]

The traditional literature analysis mainly focuses on the literature metadata such as topics, authors, keywords, references, and rarely pays attention to the main content of papers. However, in many scientific domains such as science, computing, engineering, the methods and datasets involved in the papers published carry important information and are quite useful for domain analysis and recommendation. Method and dataset entities have various forms, which are more difficult to recognize than common entities. In this paper, we propose a novel Method and Dataset Entity Recognition model (MDER), which is able to effectively extract the method and dataset entities from the main textual content of scientific papers. The model is the first to combine rule embedding, a parallel structure of Convolutional Neural Network (CNN) and a two-layer Bi-directional Long Short-Term Memory (BiLSTM) with the self-attention mechanism. We evaluate the proposed model on datasets constructed from the published papers of different research areas in computer science. Our model performs well in multiple areas and features a good capacity for cross-area learning and recognition. Ablation study indicates that building different modules collectively contributes to the good entity recognition performance as a whole. The data augmentation positively contributes to model training, making our model much more robust. We finally apply the proposed model on PAKDD papers published from 2009–2019 to mine insightful results over a long time span. 1 1 PAKDD is the abbreviation of Pacific-Asia Conference on Knowledge Discovery and Data Mining. Our source code and datasets are available at https://github.com/houlinlinvictoria/MDER. [ABSTRACT FROM AUTHOR]

In this paper, we propose a new fuzzy time series (FTS) forecasting method based on optimal partitions of intervals in the universe of discourse and optimal weighting vectors of two-factors second-order fuzzy-trend logical relationship groups (TSFTLRGs). The proposed method uses particle swarm optimization (PSO) techniques to obtain the optimal partitions of intervals and the optimal weighting vectors simultaneously. The proposed FTS forecasting method outperforms the existing methods for forecasting the TAIEX and the NTD/USD exchange rates in terms of forecasting accuracy rates. It provides us with a useful way to deal with forecasting problems to get higher forecasting accuracy rates. [ABSTRACT FROM AUTHOR]

To contain the competitive influence spread in social networks is to maximize the influence of one participant and contain the influence of its opponent. It is desirable to develop effective strategies for influence spread of the participants themselves instead of blocking the influence spread of their opponents. In this paper, we extend the linear threshold model to establish the diffusion-containment model, abbreviated as D-C model, by incorporating the realistic specialties and characteristics of the containment of competitive influence spread. Then, we discuss the influence spread mechanism for the D-C model, and give the algorithm for the propagation of the diffusion influence (D-influence) and containment influence (C-influence). Further, we define the sub-modular set function of the C-influence in the D-C model and consequently give a greedy algorithm for solving the problem of maximizing the competitive influence containment approximately. Experimental results show the feasibility of our method. [ABSTRACT FROM AUTHOR]

This paper proposes a novel technique to create scenarios that can be used by a user simulator for exhaustively evaluating spoken dialogue systems. The scenarios are automatically created from simple scenario-templates that the systems' developers create manually employing their knowledge about typical goals of the system's users. The scenarios contain goals, which the user simulator will try to achieve through the interaction with the systems. The goals are represented in the form of semantic frames, which are associated with user utterances of sentences and are taken from utterance corpora. In this way, the scenarios enable speech-based interaction between the simulator and the spoken dialogue systems to be evaluated. Experiments have been carried out employing two spoken dialogue systems (Saplen and Viajero), a user simulator and two utterance corpora previously collected for two different application domains: fast-food ordering and bus travel information. Experimental results show that the technique has been useful for exhaustively evaluating the systems and finding out problems in their performance that must addressed to improve them. Some of these problems are caused by acoustic similarity between some uttered words and strong speaker accents. Thus, we think these problems would have been difficult to uncover employing the user simulation techniques typically used nowadays, as they do not employ real speech and just consider semantics of user intentions. [ABSTRACT FROM AUTHOR]

This paper addresses the problem of evaluating the node importance in actual complex networks. Firstly, the indicators used to evaluate the node importance are defined based on complex network theory, and their characteristics are analyzed in detail. Besides, a new indicator is put forward on the basis of K-shell decomposition, named improved K-shell. Secondly, in order to evaluate the node importance comprehensively, a multi-attribute ranking method is proposed based on the Technique for Order Preference by Similarity to Ideal Object (TOPSIS). Finally, our method is used to study two actual cases. Results show that, our method outperforms other methods in distinguishing the node importance of actual complex networks and can provide scientific decision support for the administration department. [ABSTRACT FROM AUTHOR]

The growing availability of large collections of digitized artworks has disclosed new opportunities to develop intelligent systems for the automatic analysis of fine arts. Among other benefits, these tools can foster a deeper understanding of fine arts, ultimately supporting the spread of culture. However, most of the systems proposed in the literature are only based on visual features of digitized artwork images, which are sometimes only integrated with some metadata and textual comments. A Knowledge Graph (KG) that integrates a rich body of information about artworks, artists, painting schools, etc., in a unified structured framework, can provide a valuable resource for more powerful information retrieval and knowledge discovery tools in the artistic domain. To this end, in this paper we present A r t G r a p h : 1 1 A r t G r a p h and associated code are publicly available on https://doi.org/10.5281/zenodo.6337958.. an artistic KG based on WikiArt and DBpedia. The graph already provides knowledge discovery capabilities without having to train a learning system. In addition, we propose a novel KG-enabled fine art classification method based on A r t G r a p h , which is used to perform artwork attribute prediction tasks. The method extracts embeddings from A r t G r a p h and injects them as "contextual" knowledge into a Deep Learning model. Compared to the state-of-the-art, the proposed model provides encouraging results, suggesting that the exploitation of KGs in combination with Deep Learning can pave the way for bridging the gap between the Humanities and Computer Science communities. • Automatic art analysis can promote a deeper understanding of fine arts. • Knowledge Graphs and Deep Learning can provide tools to support this analysis. • We introduce A r t G r a p h , an artistic Knowledge Graph. • We present a classification method that jointly exploits visual and graph embeddings. • The approach can help bridge the gap between the Humanities and Computer Science. [ABSTRACT FROM AUTHOR]

Boolean Satisfiability problem (SAT) plays a prominent role in many domains of computer science and artificial intelligence due to its significant importance in both theory and applications. Algorithms for solving SAT problems can be categorized into two main classes: complete algorithms and incomplete algorithms (typically stochastic local search (SLS) algorithms). SLS algorithms are among the most effective for solving uniform random SAT problems, while hybrid algorithms achieved great breakthroughs for solving hard random SAT (HRS) problem recently. However, there is a lack of algorithms that can effectively solve both uniform random SAT and HRS problems. In this paper, a new SLS algorithm named SelectNTS is proposed aiming at solving both uniform random SAT and HRS problem effectively. SelectNTS is essentially an improved probability selection based local search algorithm, the core of which includes new clause and variable selection heuristics: a new clause weighting scheme and a biased random walk strategy are utilized to select a clause, while a new probability selection strategy with the variation of configuration checking strategy is used to select a variable. Extensive experimental results show that SelectNTS outperforms the state-of-the-art random SAT algorithms and hybrid algorithms in solving both uniform random SAT and HRS problems effectively. [ABSTRACT FROM AUTHOR]

Attribute-level schema matching is a critical step in numerous database applications, such as DataSpaces, Ontology Merging and Schema Integration. There exist many researches on this topic, however, they all ignore evidences about the positions of attributes in query statements, which are crucial to find high-quality matches between schema attributes. In this paper, we propose a novel matching technique based on the positions of attributes appearing in the schema structure of query results. The positions of attributes in query results embody the extent of the importance of an attribute for the user browsing the query results. The core idea of our approach is to collect the statistics about attribute positions from query logs to find correspondences between attributes ( matches ). Our method works in three phases. The first phase is to design a matrix to record the statistics about attribute positions. Then, we employ two scoring functions to measure the similarities between collected statistics of two schemas to be matched. Finally, we employ a traditional algorithm to find the optimal mapping. Furthermore, our approach can be combined with other existing matchers to obtain more accurate matching results. An experimental study shows that our approach is effective and has good performance. [ABSTRACT FROM AUTHOR]

Being aware of people’s unuttered knowledge needs is the prerequisite of providing “active” and “in time” knowledge assistance. However, such an awareness of knowledge needs has been achieved at a high cost since most existing methods rely on the manually defined rules or a large amount of user data to work. In this paper, we formulate the problematic situations in task processing as knowledge application context (KAC), and propose to elicit KACs semi-automatically from domain Q&A archives. Assuming that the KACs frequently occurring and semantically matching with the user’s task context are more likely to imply the knowledge needs of the user, we design a mechanism of knowledge need awareness (KNA) to predict users’ knowledge needs in complex tasks. Experimental results show that the proposed method has significantly outperformed the information retrieval approaches used as baselines. The study provides a new method for reusing the contextualized knowledge in Q&A and thereby opens up a new way to build efficient active knowledge systems. [ABSTRACT FROM AUTHOR]

Currently, the research in computer science has been exponentially expanded beyond its own fields into the other research fields such as medical science, business, and social science in forms of collaborative researches. This collaborative researches stimulate a new recommending algorithm for determining a potential research collaborator under the interdisciplinary environment. Unlike other research fields, the research problems in computer science can be transformed to other known and solvable problems. In this paper, a new hybrid algorithm based on dynamic collaboration over time was proposed for recommending an appropriate collaborator. Besides considering only three basic factors concerning social proximity, friendship, and complementarity skill as employed by others’, three new additional factors related to research interest, up-to-date publication data, and seniority of researcher are involved in our analysis. A set of new measures for all six recommending factors were proposed. The experiments were conducted with real bibliographic data within six continuous years of publication and over six topics in computer science. Our results were significantly higher than the results of the other methods at 90% confidence level. [ABSTRACT FROM AUTHOR]

Consider two point sets in the plane, a set of points of interest and a set of query points that is used to establish distance restrictions with respect to the set of points of interest. A nearest/farthest spatial skyline query retrieves the subset of desirable or relevant points of interest, called skyline points, such that no other point of interest is simultaneously closer to/farther from all the query points. The nearest/farthest top- k spatial skylines, are the best k nearest/farthest spatial skylines among the existent ones. All these queries find applications in decision-making support systems, facility location, crisis management and in trips or events planning. To take into account that each point of interest has a different importance, a weight is assigned to each of them and multiplicative weighted Euclidean distances are used. In this paper, we study, for the first time, the nearest and farthest spatial skyline queries when multiplicative weighted Euclidean distances are considered. We prove that most of the properties of the traditional non weighted nearest and farthest spatial skyline queries are no longer true under the weighted Euclidean distance and, consequently, the strategies used for solving non weighted spatial skyline queries are not usable in the weighted case. We present a sequential and a parallel algorithm, to be run on the CPU and on a Graphics Processing Unit, respectively, for solving nearest/farthest weighted spatial skyline queries and to extract the nearest/farthest top- k spatial skylines. We provide the time and space complexity analysis of both algorithms together with their theoretical comparison. We also have developed a simple interface to deal with weighted spatial skyline queries which allows to visualize and store in a file the obtained spatial skylines. Finally, we present and discuss experimental results obtained with the implementation of the proposed sequential and parallel algorithms. • Suggest a new measure to determine the spatial skyline points considering distance and importance. • Develop a mathematical study of the geometric properties of the weighted spatial skyline points. • Propose a sequential and a parallel algorithm to obtain all or the top-k spatial skyline points. • Algorithms are theoretically and experimentally analyzed and compared. • The experimental results prove that the parallel algorithm is robust, efficient and faster than the sequential algorithm. [ABSTRACT FROM AUTHOR]