By means of "Arc Welding Information Analysis Platform", the current progress in the research and application of arc welding process information analysis methods is introduced. It proposes how to use data to find problems and obtain new cognition in data; how to locate problems and find analytical "blind spots" according to data; and provides a way to realize the understanding of key information in the arc welding process through analysis data with multiple testing examples in welding production. That is, taking the robot production line of the digital welding production workshop as an example, the output characteristics of different welding machines can intuitively produce quality differences in the weld seam and the "short boards" in the implementation of data supervision standards; taking the process design of multilayer welding of aluminum alloy plates as an example, the detection of the transition time of the droplets in pulsed current welding, as well as the regulation of the transition amount of the droplets are discussed; the fluctuation information of the contact resistance between the welding wire and the conductive nozzle and its influence on the arc voltage drop are revealed, which reveals the corresponding arc characteristics generated by the oxidation of the surface of the aluminum alloy welding wire, which also provides a new analysis method for the surface state of the aluminum alloy welding wire and the arc stability of the welding process. The article specially introduces the mapping of "visual" information from image to data, laying the foundation for deep learning of data information and intelligent recognition of images.