Method and system for constructing visual chart based on natural language

The invention relates to a method and system for constructing a visual chart based on a natural language, and the method comprises the steps: receiving a natural language and a data attribute as input through an attribute-level model, and predicting an attribute and a filter for each attribute; the context state machine determines the content needing to be visualized according to the context, the prediction attribute of the current sentence and the filter; and constructing a visual chart through a visualization generation module on the basis of the attributes and the filters which need to be visualized. By adopting the method disclosed by the invention, the fuzzy and diversified natural language is analyzed into the data attributes and the filtering conditions based on a deep learning method, and then the visual chart is constructed for the extracted data attributes to respond to the natural language of the user, so that the threshold of data visualization is reduced, the practicability is high, and the metho