With the evolution of Internet, the meaning and accessibility of text documents and electronic information has increased. The automatic text categorization methods became essential in the information organization and data mining process. A proper classification of e-documents, various Internet information, blogs, emails and digital libraries requires application of data mining and machine learning algorithms to retrieve the desired data. The following paper describes the most important techniques and methodologies used for the text classification. Advantages and effectiveness of contemporary algorithms are compared and their most notable applications presented.
1 September 2016 in International Journal of Applied Mathematics and Computer Science
The aim of this paper is to introduce a strategy to find a minimal set of test nodes for diagnostics of complex analog systems with single parametric faults using the support vector machine (SVM) classifier as a fault locator. The results of diagnostics of a video amplifier and a low-pass filter using tabu search along with genetic algorithms (GAs) as node selectors in conjunction with the SVM fault classifier are presented. General principles of the diagnostic procedure are first introduced, and then the proposed approach is discussed in detail. Diagnostic results confirm the usefulness of the method and its computational requirements. Conclusions on its wider applicability are provided as well.
1 November 2010 in International Journal of Electronics and Telecommunications
The aim of this paper is to present the analysis and modeling of the hysteresis phenomenon using a Preisach operator. The fundamentals of parameterized hysteresis modeling are introduced, by utilizing three probability density functions. Then, the Preisach operator and its characteristics are defined. Subsequently, results of simulations obtained by means of the aforementioned functions are presented and compared to the ones obtained by other authors.
11 July 2025 in International Journal of Electronics and Telecommunications
The following paper presents research on the Automatic Speech Recognition (ASR) methods for the construction of a system to automatically transcribe the medical interview in Polish language during a visit in the clinic. Performance of four ASR models based on Deep Neural Networks (DNN) was evaluated. The applied structures included XLSR-53 large, Quartznet15x5, FastConformer Hybrid Transducer-CTC and Whisper large. The study was conducted on a self-developed speech dataset. Models were evaluated using Word Error Rate (WER), Character Error Rate (CER), Match Error Rate (MER), Word Accuracy (WAcc), Word Information Preserved (WIP), Word Information Lost (WIL), Levenshtein distance, Jaro - Winkler similarity and Jaccard index. The results show that the Whisper model outperformed other tested solutions in the vast majority of the conducted tests. Whisper achieved a WER = 20.84%, where XLSR-53 WER = 67.96%, Quartznet15x5 WER = 76.25%, FastConformer WER = 46.30%. These results show that Whisper needs further adaptation for medical conversations, as current volume of transcription errors is not practically acceptable (too many mistakes in the description of the patient's health description).
The paper presents the automated method of the soil profile generation using the unsupervised learning. This method can be used as the main source of knowledge for the future soil classification module, which would identify geotechnical layers without the interaction with the user. The method uses the ant clustering, which groups measurements from the Dilatometer of Marchetti Test and Cone Penetration Test. The test site, algorithm and its application to the generation of soil profiles in the Warsaw University of Life Sciences campus are described. Results are discussed to provide a more general insight into application of intelligent algorithms into the geotechnical data analysis.
25 November 2024 in International Journal of Electronics and Telecommunications
The paper presents the methodology for evaluating mobile game players retention, which is the basis for generating economic income for its creators. The system for collecting and processing the in-game data (events related to the players’ actions) exploiting the Big Data cloud platform is described. The player profile with the crucial features allowing for the retention analysis is introduced. Datasets generated for the My Spa Resort mobile game by CherryPick company are described. The retention prediction approach based on the similarity estimation between the analyzed and already inactive players is presented. Results of the prediction using the k Nearest Neighbors (kNN) classifier are discussed.