CPC G06F 16/958 (2019.01) [G06F 16/951 (2019.01); G06F 16/953 (2019.01); G06F 40/30 (2020.01); G06V 20/41 (2022.01)] | 25 Claims |
1. A method comprising:
obtaining a digital content the digital content comprises multiple visually separated components, wherein the multiple visually separated components comprise a component and a second component
automatically analyzing at least a portion of the digital content to extract features associated with the component;
automatically determining, based on the extracted features, a semantic meaning of the component, wherein said automatically determining is performed without using manually inputted hints in the digital content, wherein said automatically determining comprises:
determining for the component, a set of candidate semantic meanings related to the component and respective likelihoods of the component having each candidate semantic meaning, wherein each respective likelihood is indicative of a score measuring a probability that the component has the candidate semantic meaning, wherein the set of candidate semantic meanings comprises the semantic meaning having a first likelihood and a second semantic meaning having a second likelihood, wherein the first likelihood is lower than the second likelihood, wherein the second semantic meaning is different than the semantic meaning;
determining for the second component a third likelihood of the second component having the second semantic meaning, wherein the third likelihood is higher than the second likelihood;
determining, based on the extracted features, the second semantic meaning of the second component; and
based on the second likelihood and the third likelihood, determining that the component has the semantic meaning; and
automatically and without user intervention, performing an action associated with the digital content, wherein the action is determined based on the semantic meaning of the component, wherein said method is performed without utilizing manual hints provided by a user in the digital content.
|