Domain knowledge based feature extraction for enhanced text representation

A system for representing text, the system comprising:

• a memory;
• and
• a processor coupled to the memory, wherein the processor performs operations, the operations performed by the processor comprising:
• receiving a text;
• analyzing the text by utilizing a pre-trained embedding model and a feature vector model, wherein selected words in the text have corresponding weights;
• performing operations whose parameters include weights of a feature vector and an embedding to generate a weighted embedding data structure, wherein the feature vector includes a first plurality of keywords that comprise words of a sentence, and at least a one additional keyword that provides additional information, wherein the one additional keyword is not in the sentence, and wherein each of the first plurality of keywords and the one additional keyword have associated weights that correspond to the weights of the feature vector;
• performing a summation of all corresponding columns of a plurality of rows of the weighted embedding data structure to generate a data structure that represents the text;
• and
• utilizing the data structure that represents the text to generate at least one of a classification metadata for the text and a summarization of the text.

A computer program product for representing text, the computer program product comprising a computer readable storage medium having computer readable program code embodied therewith, the computer readable program code configured to perform operations, the operations comprising:

• receiving a text;
• analyzing the text by utilizing a pre-trained embedding model and a feature vector model, wherein selected words in the text have corresponding weights;
• performing operations whose parameters include weights of a feature vector and an embedding to generate a weighted embedding data structure, wherein the feature vector includes a first plurality of keywords that comprise words of a sentence, and at least a one additional keyword that provides additional information, wherein the one additional keyword is not in the sentence, and wherein each of the first plurality of keywords and the one additional keyword have associated weights that correspond to the weights of the feature vector;
• performing a summation of all corresponding columns of a plurality of rows of the weighted embedding data structure to generate a data structure that represents the text;
• and
• utilizing the data structure that represents the text to generate at least one of a classification metadata for the text and a summarization of the text.