Presentation materials
In modern conditions of rapid growth of textual information volumes, efficient extraction of named entities becomes a key aspect of data analysis in various fields of science and technology.
The task of text data analysis is extremely relevant for a number of internal services of the Joint Institute for Nuclear Research (JINR), in particular, in the context of development of the JINRex...
The integration of different high-energy collision Monte Carlo models into a unified simulation process is inherently time-consuming, largely due to the fact that they are typically developed as monolithic applications. Diverse data formats of the aforementioned models often necessitate the use of numerous converters and supplementary scripts, which can significantly impede the modelling...
Baikal-GVD is a neutrino telescope with an effective volume of about 1 km^3 located in Lake Baikal. To enable observations within the framework of neutrino astronomy, the following event processing problems must be solved in the experiment:
1) selection of the neutrino (ฮฝ) component against the background of events, caused by extensive air showers (EAS);
2) reconstruction of the parameters...
Conducting research on the morphofunctional state of the central nervous system of small laboratory animals, it is necessary to create a convenient environment for data analysis, including one that allows automating the routine stages of their processing. ะกonducting behavioral experiments, researchers face the problem of incorrect detection of an animal in a behavioral test system or maze when...
Event indexing, or event metadata systems (EMS) are common for particle physics experiments. Their main goal is to keep a searchable catalogue of experimental events from which a subset of data can be extracted based on given filtering criteria. The BM@N experiment's EMS has been designed, developed and deployed previously and is now being improved to increase its performance, convenience for...
In the course of developing a software ecosystem of the BM@N experiment, a multitude of web services are created, including data processing systems and information resources. A significant challenge is the monitoring of system operational states as an integral component of the maintenance procedure. The report considers implementing a contemporary solution for log management of the BM@N...
The BM@N experiment, as part of the NICA complex, produces a substantial quantity of physics data, necessitating the implementation of a sophisticated infrastructure for the efficient storage, processing, and management of the data. In order to address these challenges, a comprehensive set of information systems has been developed. The complex includes an information system representing the...
The BM@N 8th physics run using Xenon ion beams was successfully completed in February 2023, resulting in the recording of approximately 550 million events. These events were recorded in the form of aproximately 30000 files, with a combined size exceeding 400 TB. The processing of the BM@N files is done in two steps: converting files from Raw format to Digi and on the next step converting from...
One of the principal technical characteristics of the SPD (Spin Physics Detector) is its triggerless data acquisition. The data acquisition system (DAQ) aggregates data from the detectors of the facility and organizes them into blocks for further primary processing. This approach allows for data arrival rates of up to 20 Gb/sec, with the annual volume of collected data reaching hundreds of...
The first experience of using Rucio to manage SPD data
Modern experiments in high energy physics are characterized by the need to process huge amounts of data.
SPD (Spin Physics Detector) is a universal detector of the NICA collider (Nucleotronโbased Ion Collider fAcility), being built at the Joint Institute for Nuclear Research (Dubna), and designed to study the spin structure of...
Neutron tomography is a powerful tool in material science due to the large penetration depth, sensitivity to light elements and good contrast for elements with close atomic numbers. Such properties of the interaction of neutrons with matter make it possible to obtain data on the internal structure of objects that complement X-ray tomography. However, a significant disadvantage of the method...
A proprietary method for finding the global minimum to optimise the loss function of a neural network has been developed. The method involves the use of an arbitrary loss function. In this study, a function based on the calculation of the area under the ROC curve (AUC) was implemented. The application of this method in combination with the optimisation of the hyperparameters of the neural...
Collaborative work on documents is an essential part of scientific research, especially in large collaborations. Various software tools are used to organize and facilitate this process as well as to store files. Some of the most widely used tools in the scientific community include DocDB, XWiki, and the CERN Document Server. Currently, several independent instances of the DocDB system have...
The heterogeneous HybriLIT platform is used at JINR to solve various computational tasks, including data collection and processing from experiments, as well as modeling and simulation of physical processes. The main component of the platform is the Govorun supercomputer, which consists of more than a hundred servers that constantly run user programs and communicate with each other over the...
Monitoring of computing cluster resources provides detailed information about the status of various components of computing nodes in real time. Such a tool helps to monitor the status of CPU, RAM, software-defined storage, etc. only at the current moment. To evaluate the efficiency of a computing cluster, it is necessary to analyze the statistics of the use of various components of computing...
JINR Institutional repository is built on DSPACE software platform. Repository is a hub of results made by JINR employees, showcase of scientific activities of the Institute.
At the first stage we store publication automatically imported via API from eLIBRARY and INSPIREHEP databases from 2019-2024 yy and then we will continue to go feather into the history. The repository allows users to...
Nowadays, the successful implementation of a significant part of scientific projects involves the use of a Distributed Information and Computing Environment (DICE) for storing, processing and analyzing data. The JINR DICE initiative is dedicated to the creation, support and development of such an environment by combining the resources of educational and research organizations of the JINR...
