Processing folder extracting and summarizing content

A computer-implemented method that utilizes a processor that executes computer executable components stored in memory to perform the following acts:

• receiving input from a user pertaining to a subject of interest;
• obtaining media files pertaining to the subject of interest;
• populating a knowledge base with the obtained media files;
• determining a type of media for each of the obtained media files;
• extracting relevant features and metadata from each of the obtained media files, based on the determined media type of each of the obtained media files, wherein extracting the relevant features and metadata comprises utilizing media-type-specific feature extraction pipelines including extracting bag-of-words representations or sentiment analysis for text documents, performing object detection or classification for image files, or performing speech recognition or audio classification for audio files;
• utilizing a large language model to extract key information from the obtained media files and to generate embeddings representing content of the obtained media files;
• storing the embeddings in a vector database configured to enable comparison of the obtained media files based on the embeddings;
• categorizing and tagging the obtained media files based on complexity, length, or target audience of the obtained media files;
• generating a summary of each of the obtained media files based on the extracted key information;
• determining relationships between the obtained media files based on the extracted key information and the embeddings;
• determining a chronological order of the obtained media files to be reviewed by a user, based on the generated summary, the extracted key information, and the determined relationships;
• generating a graph displaying the determined chronological order;
• receiving feedback from a user;
• and
• updating the determined chronological order and graph, based on the received feedback.

A computer program product comprising a computer readable storage medium having program instructions embodied therewith, the program instructions executable by a processor to cause the processor to:

• receive input from a user pertaining to a subject of interest;
• obtain media files pertaining to the subject of interest;
• populate a knowledge base with the obtained media files;
• determine a type of media for each of the obtained media files;
• extract relevant features and metadata from each of the obtained media files, based on the determined media type of the each of the obtained media files, wherein extracting the relevant features and metadata comprises utilizing media-type-specific feature extraction pipelines including extracting bag-of-words representations or sentiment analysis for text documents, performing object detection or classification for image files, or performing speech recognition or audio classification for audio files;
• utilize a large language model to extract key information from the obtained media files, and to generate embeddings representing content of the obtained media files;
• store the embeddings in a vector database configured to enable comparison of the obtained media files based on the embeddings;
• categorize and tag the obtained media files based on complexity, length, or target audience of the obtained media files;
• summarize each of the obtained media files based on the extracted key information;
• determine relationships between the obtained media files based on the extracted key information and the embeddings;
• determine a chronological order of the obtained media files to be reviewed by a user, based on the generated summary, the extracted key information, and the determined relationships;
• generate a graph displaying the determined chronological order;
• receive feedback from a user;
• and
• update the determined chronological order and graph, based on the received feedback.