Analysis of the sentence writing test is conducted in this paper to support diagnostics of the Parkinsons disease. Drawing and writing tests digitization has become a trend where synergy of machine learning techniques on the one side and knowledge base of the neurology and psychiatry on the other side leading sophisticated result in computer aided diagnostics. Such rapid progress has a drawback. In many cases, decisions made by machine learning algorithm are difficult to explain in a language human practitioner familiar with. The method proposed in this paper employs unsupervised learning techniques to segment the sentence into the individual characters. Then, feature engineering process is applied to describe writing of each letter using a set of kinematic and pressure parameters. Following feature selection process applicability of different machine learning classifiers is evaluated. To guarantee that achieved results may be interpreted by human, two major guidelines are established. The first one is to keep dimensionality of the feature set low. The second one is clear physical meaning of the features describing the writing process. Features describing amount and smoothness of the motion observed during the writing alongside with letter size are considered. Resulting algorithm does not take into account any semantic information or language particularities and therefore may be easily adopted to any language based on Latin or Cyrillic alphabets.

This paper presents the use of gamified crowdsourcing for knowledge content validation. Constructing a high-quality knowledge base is crucial for building an intelligent system. We develop a refinement process for the knowledge base of our word retrieval assistant system, where each piece of knowledge is represented as a triple. To validate triples acquired from various sources, we introduce yes/no quizzes and present them to many casual users for their inputs. Only the triples voted “yes” by a sufficient number of users are incorporated into the main knowledge base. Users are incentivized by rewards based on their contribution to the validation process. To ensure transparency of the reward-giving process, blockchain is utilized to store logs of the users’ inputs from which the rewards are calculated. Different strategies are also proposed for selecting the next quiz. The simulation results indicate that the proposed approach has the potential to validate knowledge contents. This paper is a revised version of our conference paper presented at the 12th Asian Conference on Intelligent Information and Database Systems (ACIIDS 2020).

This paper addresses the issue of optimal allocation of spare modules in large series-redundant systems in order to obtain a required reliability under cost constraints. Both cases of active and standby redundancy are considered. Moreover, for a subsystem with standby redundancy, two cases are examined: in the first case, all the spares are maintained in cold state (cold standby redundancy) and, in the second one, to reduce the time needed to put a spare into operation when the active one fails, one of the spares is maintained in warm conditions. To solve this optimization problem, for the simpler case of active redundancy an analytical method based on the Lagrange multipliers technique is first applied. Then the results are improved by using Pairwise Hill Climbing, an original fine-tuning algorithm. An alternative approach is an innovative evolutionary algorithm, RELIVE, in which an individual lives for several generations and improves its fitness based on local search. These methods are especially needed in case of very large systems.

Predicting the solutions of complex systems is a crucial challenge. Complexity exists because of the uncertainty as well as nonlinearity. The nonlinearity in complex systems makes uncertainty irreducible in several cases. In this paper, two new approaches based on neural networks are proposed in order to find the estimated solutions of the fully fuzzy nonlinear system (FFNS). For obtaining the estimated solutions, a gradient descent algorithm is proposed in order to train the proposed networks. An example is proposed in order to show the efficiency of the considered approaches.

Efficient operation of the integrated management information systems (IMISs), especially multi-agent systems, is related to their ability to automatically process collective knowledge. On the basis of this knowledge the decision-making process is realized in the business organizations. This paper presents issues related to framework for acquiring and acquisition subprocesses in a collective knowledge of business organization processing in IMIS. The main novelty of the developed framework is the coverage of all the areas of operation of an organization. Additionally, the inter-area knowledge for automatic strategic-level decision-making has been taken into consideration. The main improvements of this framework are that it allows for processing of the whole collective knowledge of business organization and it can be directly implemented in IMIS.

In this paper, we propose an evacuation support system that provides evacuation routes in case of disasters, and verify the usefulness of the system. Current popular wireless communication infrastructure is supported by a series of base stations and one base station handles a lot of communication. Therefore, when communication base stations break down due to disasters such as an earthquake, it may become difficult for people to use their smartphones based on the Internet. When the communication infrastructure is paralyzed, people will have great difficulty collecting information about the conditions of transportation systems and about the safety of family and friends using smartphones. Our proposed system addresses this problem by using multiple kinds of mobile agents in addition to static agents on smartphones that use a mobile ad hoc network (MANET). The proposed system collects information with mobile agents and diffuses information via mobile agents so that the system is able to provide an optimized evacuation route for each user in a dynamically changing disaster environment. In this paper, we extend our previously proposed evacuation support system to consider elevation information when constructing evacuation routes. When a tsunami or a flood tide occurs, low elevations may be under water. Therefore, this revised evacuation support system helps people to move to safer places by selecting higher elevation routes when warranted by the situation.

Mining the valuable knowledge from real data has been a hot topic for a long time. Repeating pattern is one of the important knowledge, occurring in many real applications such as musical data and medical data. In this paper, our purposes are to contribute an efficient mining algorithm for repeating patterns and to conduct a real application using the repeating patterns mined. In terms of mining the repeating patterns, although a number of past studies were made on this issue, the performance cannot still earn the users’ satisfactions especially for large data sets. For this issue, in this paper, we propose an efficient algorithm named Fast Mining of Repeating Patterns, which achieves high performance of discovering the repeating patterns by a novel index called Quick-Pattern Index. In terms of applications, a music recommender system named repeating-pattern-based music recommender system is proposed to deal with problems in music recommendation. Even facing a very sparse rating matrix, the recommendation can still be completed. The experimental results show that our proposed mining algorithm and recommender system outperform the previous works in terms of efficiency and effectiveness, respectively.

Credit scoring is a crucial problem in both finance and banking. In this paper, we tackle credit scoring as a classification problem where three local search-based methods are studied for feature selection. The feature selection is an interesting technique that can be launched before the data classification task. It permits to keep only the relevant variables and eliminate the redundant ones which enhances the classification accuracy. We study the local search method (LS), the stochastic local search method (SLS) and the variable neighborhood search method (VNS) for feature selection. Then, we combine these methods with the support vector machine (SVM) classifier to find the best described model from a dataset with the correct class variable. The proposed methods (LS+SVM, SLS+SVM and VNS+SVM) are evaluated on both German and Australian credit datasets and compared with some well-known classifiers. The numerical results are promising and show a good performance in favor of our methods.

It is my pleasure to present

Bound to 140 characters, tweets are short and ambiguous by nature. It can be hard for a user without any kind of context to effectively understand what the tweet is about. Due to this restriction, it is, therefore, necessary to know the tweet’s context to make it easily understandable to a reader. In this paper, we treat the problem of tweet contextualization. We propose a specific method allowing to automatically contextualize tweets using information coming from social user interactions. Contrary to classical contextualization methods that only consider text information which is insufficient, since text information on Twitter is very sparse, we combine different types of signals (social, temporal, textual). Our experimental results validate the benefits of our approach and confirm that generated contexts contain relevant information with given tweet.