| CPC G06F 16/358 (2019.01) [G06F 40/284 (2020.01); G06F 40/30 (2020.01); G06F 40/40 (2020.01); G06V 10/454 (2022.01)] | 20 Claims |
|
1. A computer implemented method comprising:
executing, by a content classifier, a querying process that searches for and returns a first database document and a second database document each containing natural language text associated with a database environment;
tokenizing, by the content classifier, the first and second database documents into a series of n-gram tokens;
grouping, by the content classifier, the n-gram tokens into a plurality of topic classes using natural language processing (NLP);
generating, by a structure classifier, a feature map by applying a convolution layer to an image that depicts a database structure graph of the database environment, wherein nodes of the database structure graph represent architectural elements of the database environment and edges of the database structure graph represent associations between respective pairs of architectural elements;
detecting, by the structure classifier, an architectural element in the image by applying a region proposal network (RPN) to the feature map;
classifying, by the structure classifier, the architectural element into one of the plurality of topic classes using a trained neural network; and
rendering, responsive to detecting a user selection of the architectural element via a graphical user interface that depicts at least a portion of the database structure graph, an overlay depicting information about the architectural element using text and other architectural elements that are in a same topic class as the user-selected architectural element.
|