In the current digital world, data plays a crucial role and is necessary for working with different systems, is important for decision support systems, as well as for providing high effectiveness of different services for society. They are collected in different areas of life and human activity. They come from the world around us; they are generated by various systems, including land and satellite systems. On the other hand, it is not important to have the data only, but the skilful and effective use of it, for example, using a machine learning paradigm or in the case of natural language processing (NLP). Machine learning is an example where data is needed, and the machine learning tools are a core of the current intelligent and advanced systems, and they have to deal with this data, process it, it also means they must be able to take it. In the case of NLP without data preprocessing, it wouldn't convert complex, often unmeasurable information into numerical vectors in a multidimensional space. Data processing is also important for deep learning, when involves hierarchical feature extraction and identifying complex abstract concepts (objects) to the final image recognition or speech understanding. The data are also a part of analytics and reporting levels in business, medicine, or in commercial and production spaces. Thus, the pre-processing is also very important to obtain the desired effect and to be able to analyse these large data at all.
This special session is dedicated to data prep strategies, in context, various uses of them, including, for example, machine learning. The session is also open for different problems that exist in the domain of embedding, as well as, deep learning. Strategies of working with data may be dedicated to a number of activities or processes, from data cleansing, noise detection and elimination, data editing, instance selection, noise reduction, eliminate the outliers and detecting wrong or distorted labels, resampling, feature extraction and selection, to data transformation. Data preparation can be also merged with an elimination of the class imbalance problem, data blending, aggregate data from diverse sources and with data wrangling. Data prep strategies can be merged with processes relatively easy or complex, and can be based on simple tools or require complex computations. It can also be merged with novelty approaches for embedding. It is also possible that these strategies require the solution of problems that belong to the group of optimisation problems, with the aim of eliminating barriers hidden in the data for the further application of analysis tools. So, it is a reason. that the data prep strategies can also be merged with optimisation tools.
The topics of interest for this session include, but are not limited to:
All contributions must be of high quality, original, and must not have been previously published elsewhere or intended for publication elsewhere.
The papers will be reviewed by the International Program Committee. The best submissions will be selected for presentation and will be included in the conference proceedings.
The conference proceedings will be published in Elsevier's Procedia Computer Science open access journal, available in ScienceDirect and submitted to be indexed/abstracted in CPCI (ISI conferences and part of Web of Science), Engineering Index, and Scopus.
Authors of selected papers may be invited to submit extended versions of their papers for publication as full journal papers, for example in the KES Journal or other journals.
Submitted papers should be prepared in the Procedia style and should be limited to 10 pages (minimum 8 pages). All papers must be submitted electronically through the EasyChair online submission and review system.
.
Guidance notes for the preparation of paper is available ..here...
Ireneusz Czarnowski, Gdynia Maritime University, Poland - i.czarnowski@umg.edu.pl
Antonio J. Tallón-Ballesteros, University of Huelva, Spain - antonio.tallon.diesia@zimbra.uhu.es