Vol. 40 · No. 2

Language learning applications define exercises that are pedagogical tools to introduce new language concepts. The development of this type of applications is complex due to the diversity of language learning methodologies, the variety of execution environments and the number of different technologies that can be used. This article proposes a conceptual model to develop the activities of language learning applications. It defines a new abstraction mechanism to model these activities as part of a model-driven approach to develop applications supporting different language learning processes running on different hardware and software platforms. We define a metamodel that describes the entities and relationships representing language learning activities as well as a series of examples that use the proposed abstraction mechanism to represent different language learning activities. The modelling process is simplified using a common representation that does not affect neither the visual presentation, nor the interaction of each activity. The article includes an evaluation that analyses the product correctness, robustness, extensibility, and reusability of the obtained code. These results conclude that the code generated using the proposed approach overcomes the code generated following a traditional approach.

Automatic speech recognition (ASR) technologies enable humans to communicate with computers. Isolated word recognition (IWR) is an important part of many known ASR systems. Minimizing the word error rate in cases of incremental learning is a unique challenge for developing an on-line ASR system. This paper focuses on on-line IWR using a recursive hidden Markov model (HMM) multivariate parameter estimation algorithm. The maximum likelihood method was used to estimate the unknown parameters of the model, and an algorithm for the adapted recursive EM algorithm for HMMs parameter estimation was derived. The resulting recursive EM algorithm is unique among its counterparts because of state transition probabilities calculation. It obtains more accurate parameter estimates compared to other algorithms of this type. In our experiment, the algorithm was implemented and adapted to several datasets for IWR. Thus, the recognition rate and algorithm convergence results are discussed in this work.

In this paper, we propose a logistic regression classification method based on the integration of a statistical learning model with linearized kernel pre-processing. The single Gaussian kernel and fusion of Gaussian and cosine kernels are adopted for linearized kernel pre-processing respectively. The adopted statistical learning models are the generalized linear model and the generalized additive model. Using a generalized linear model, the elastic net regularization is adopted to explore the grouping effect of the linearized kernel feature space. Using a generalized additive model, an overlap group-lasso penalty is used to fit the sparse generalized additive functions within the linearized kernel feature space. Experiment results on the Extended Yale-B face database and AR face database demonstrate the effectiveness of the proposed method. The improved solution is also efficiently obtained using our method on the classification of spectra data.

Extracting the facts, namely entities and relations, from unstructured sources is an essential step in any knowledge base construction. At the same time, it is also necessary to ensure the completeness of the knowledge base by incrementally extracting the new facts from various sources. To date, the knowledge base completion is studied as a problem of knowledge refinement where the missing facts are inferred by reasoning about the information already present in the knowledge base. However, facts missed while extracting the information from multilingual sources are ignored. Hence, this work proposed a generic framework for knowledge base completion to enrich a knowledge base of crime-related facts extracted from online news articles in the English language, with the facts extracted from low resourced Indian language Hindi news articles. Using the framework, information from any low-resourced language news articles can be extracted without using language-specific tools like POS tags and using an appropriate machine translation tool. To achieve this, a clustering algorithm is proposed, which explores the redundancy among the bilingual collection of news articles by representing the clusters with knowledge base facts unlike the existing Bag of Words representation. From each cluster, the facts extracted from English language articles are bootstrapped to extract the facts from comparable Hindi language articles. This way of bootstrapping within the cluster helps to identify the sentences from a low-resourced language that are enriched with new information related to the facts extracted from a high-resourced language like English. The empirical result shows that the proposed clustering algorithm produced more accurate and high-quality clusters for monolingual and cross-lingual facts, respectively. Experiments also proved that the proposed framework achieves a high recall rate in extracting the new facts from Hindi news articles.

Many information systems have a possibility to record their execution, and, in this way, to generate a trace about events describing the real system behaviour. From behaviour example records in traces of the event log, the α-algorithm automatically generates a process model that belongs to a subclass of Petri nets, known as workflow nets. One of the basic limiting assumptions of α-algorithm is that the event log needs to be complete. As a result of attempting to overcome the problem of completeness of the event log, we introduced the notion of weakly complete event logs, from which our modified technique and algorithm can produce the same result as the α-algorithm from complete logs on parallel processes. Thereby weakly complete logs can be significantly smaller than complete logs, considering the number of traces they consist of. Weakly complete logs were used for the realization of our idea of interactive parallel business process model generation.

Some attribute implications in an implicational base of a derived context of many-valued context can be inferred from some other attribute implications together with its scales. The scales are interpretation of some values in the many-valued context therefore they are a prior or an existing knowledge. In knowledge discovery, the such attribute implications are redundant and cannot be considered as new knowledge. Therefore the attribute implicational should be eliminated. This paper shows that the redundancy problem exists and formalizes a model to check the redundancy.

In recent years, the task of human pose estimation has become increasingly important, due to the large scale of usage, including VR applications, as well as higher-level tasks, such as human behavior understanding. In this paper, we introduce a novel two-stage deep learning approach named Segmentation-Guided Pose Estimation (SGPE). The pipeline is based on two neural networks working in a sequential fashion, while both models effectively process unorganized point clouds on the input. First, the segmentation network performs a pointwise classification into the corresponding body regions. In the next step, the point cloud with the per-point region assignment, forming the fourth input channel, is passed to the regression network. This way, both local and global features of the point cloud are preserved, helping the model fully maintain the body pose structure. Our strategy achieves competitive results on all of the examined benchmark datasets, and outperforms state-of-the-art methods.

Electric vehicles (EVs) are environmentally friendly and are considered to be a promising approach toward a green transportation infrastructure with lower greenhouse gas emissions. However, the limited driving range of EVs demands a strategic allocation of charging facilities, hence providing recharging opportunities that help reduce EV owners' anxiety about their vehicles' range. In this paper, we study a set covering method where self-avoiding walks are utilized to find the most significant locations for charging stations. In the corresponding optimization problem, we derive a lower bound of the number of charging stations in a transportation network to obtain full coverage of the most probable routes. The proposed method is applied to a transportation network of the southern part of Sweden.

Verification of the geographic location of a moving device is vital. This verification is important in terms of ensuring that the flying systems moving in the swarm are in orbit and that they are able to task completion and manage their energy efficiency. Cyber-attacks on unmanned aerial vehicles (UAV) in a swarm can affect their position and cause various damages. In order to avoid this challenge, it is necessary to share with each other the positions of UAV in the swarm and to increase their accuracy. In this study, it is aimed to increase position accuracy and data integrity of UAV by using blockchain technology in swarm. Experiments were conducted on a virtual UAV network (UAVNet). Successful results were obtained from this proposed study.

Enterprise information systems (EIS) play an important role in business process management. Process mining techniques that can mine a large number of event logs generated in EIS become a very hot topic. There always exist some deviations between a process model of EIS and event logs. Therefore, a process model needs to be repaired. For the process model with selection structures, the mining accuracy of the existing methods is reduced because of the additional self-loops and invisible transitions. In this paper, a method for repairing Logical-Petri-nets-based process models with selection structures is proposed. According to the relationship between the input and output places of a sub-model, the deviation position is determined by a token replay method. Then, some algorithms are designed to repair the process models based on logical Petri nets. Finally, the effectiveness of the proposed method is illustrated by some experiments, and the proposed method has relatively high fitness and precision compared with its peers.