Big Data Visualization Tools

Nikos Bikakis

DOI: https://doi.org/10.1007/978-3-319-63962-8_109-2

2023-11-19

Abstract:Data visualization and analytics are nowadays one of the corner-stones of Data Science, turning the abundance of Big Data being produced through modern systems into actionable knowledge. Indeed, the Big Data era has realized the availability of voluminous datasets that are dynamic, noisy and heterogeneous in nature. Transforming a data-curious user into someone who can access and analyze that data is even more burdensome now for a great number of users with little or no support and expertise on the data processing part. Thus, the area of data visualization and analysis has gained great attention recently, calling for joint action from different research areas and communities such as information visualization, data management and mining, human-computer interaction, and computer graphics. This article presents the limitations of traditional visualization systems in the Big Data era. Additionally, it discusses the major prerequisites and challenges that should be addressed by modern visualization systems. Finally, the state-of-the-art methods that have been developed in the context of the Big Data visualization and analytics are presented, considering methods from the Data Management and Mining, Information Visualization and Human-Computer Interaction communities

Databases,Graphics,Human-Computer Interaction

What problem does this paper attempt to address?

The paper primarily explores the challenges and solutions faced by big data visualization tools in the current era of data science. Specifically:

1. **Limitations of Traditional Visualization Systems**: Traditional visualization systems typically can only handle small-scale datasets or pre-processed data, and are unable to cope with dynamic, heterogeneous, and massive big datasets.

2. **Real-time Interaction Requirements**: Modern visualization systems need to support real-time interaction with large-scale datasets and maintain a response time within 1 second.

3. **Instant Visualization**: For large and constantly changing raw data (unprocessed), there is a need to provide support for instant visualization.

4. **Visual Scalability**: To avoid information overload (i.e., overlapping plots), effective data abstraction mechanisms are needed to handle large amounts of visual information.

5. **User Assistance and Personalized Customization**: Users are encouraged to understand and customize exploration scenarios based on their analysis needs, meeting the preferences and requirements of different users.

To address these challenges, the paper also introduces several cutting-edge methods and technologies, including data dimensionality reduction techniques, hierarchical data exploration, progressive data visualization, adaptive indexing, and in-situ data management. Additionally, the paper emphasizes several future research directions in data visualization, such as better understanding user needs, improving system performance, and developing new interactive machine learning methods. In summary, the paper aims to promote further development in the field of big data visualization by summarizing existing technologies and proposing future research directions.