Information and Data Fusion

- Overview

Data comprises raw, unprocessed facts that need context to become useful, while information is data that has been processed, organized, and interpreted to add meaning and value. This explanation sets the stage for how businesses can transform data into strategic assets through effective knowledge management.

[Saitama Prefecture, Japan - Civil Engineering Discoveries]

- The Data Fusion Process

Data fusion is the process of integrating multiple data sources to produce information that is more consistent, accurate and useful than that provided by any single source. The concept of data fusion originated from the evolved ability of humans and animals to integrate information from multiple senses to improve their survival. For example, a combination of sight, touch, smell and taste can indicate whether a substance is edible.

The data fusion process is usually classified as low, medium, or high, depending on the processing stage in which fusion occurs. Low-level data fusion combines multiple raw data sources to produce new raw data. The fused data is expected to be more informative and comprehensive than the raw input. For example, sensor fusion, also known as (multi-sensor) data fusion, is a subset of information fusion.

Data fusion is the joint analysis of multiple interrelated datasets that provide complementary views of the same phenomenon. The process of correlating and fusing information from multiple sources is often more accurate than the inferences that can be drawn from analyzing a single data set. Data fusion is a multifaceted concept with clear advantages, but at the same time there are many challenges that need to be carefully addressed.

Data fusion refers to the collection of different kinds of information into a process that produces a single model. Among the different ways to combine data from different sources, the multi-block (or multi-table) approach is a relevant choice. These methods can be used to combine different data matrices obtained using different analysis techniques.

- Multi-source Information Fusion

Multi-source information fusion is a complex estimation process that allows users to more accurately assess complex situations by effectively combining core evidence from massive, diverse and sometimes conflicting data received from multiple sources. It involves integrating information from these multiple sources to produce a specific and comprehensive unified estimate of an entity, activity or event. 

This definition has some key operational words: concrete, comprehensive and substantial. From an information-theoretical perspective, fusion, as an efficient information-processing function, must (at least ideally) increase the specificity and comprehensiveness of our understanding of entities, otherwise perform functions. 

Multi-source information fusion is sometimes implemented as a fully automated process or as a human-in-the-loop/in-the-loop process for analysis and/or decision support.

​