Annotating content with contextually relevant comments

1. A computer-implemented method for automatically augmenting online content with contextually relevant comments, the method comprising:

receiving, by a computer from a network server that hosts an online content provider, online content from the online content provider;

receiving, by the computer from the online content provider, user-entered comments associated with the online content:

classifying the comments, by a trained machine learning model on the computer, as either chit-chat if the comment corresponds to training data characterized as meaningless, unhelpful, or unimportant, or informative if the comment corresponds to training data characterized as containing valuable information:

for each comment classified as informative:

determining, by the computer, a portion of the online content to which the comment is most relevant, wherein the determining further comprises:

converting the online content data to text;

assigning, by the computer, scores to segments of the text indicating a relevance of the comment to each segment, and identifying a segment with the highest score; and

associating, by the computer, the comment with a position in the online content that corresponds to the determined portion of the online content;

selecting, by the computer, a subset of the informative comments for presentation, based on a predefined criteria, wherein the predefined criteria comprises a selection based on one or more of informative comments having a score exceeding a predefined threshold, recentness, randomness, a known or inferred profile of a user, a probabilistic basis according to a predefined distribution, or in real time, concurrently with user interaction with the online content;

converting the selected subset of the informative commands from text to a format suitable for presentation with the online content data; and

annotating, by the computer, the online content with the selected subset of informative comments at the positions in the online content associated with respective comments in the subset of informative comments.

2. A method in accordance with claim 1 , wherein the online content is audio or video data, and the method further comprises:

converting, by the computer, speech in the online content to text; and wherein determining comprises determining, by the computer, a portion of the text to which the comment is most relevant.

3. A method in accordance with claim 2 , wherein determining a portion of the text to which the comment is most relevant comprises assigning, by the computer, scores to segments of the text indicating a relevance of the comment to each segment, and identifying a segment with the highest score.

4. A method in accordance with claim 3 , wherein assigning, by the computer, scores to segments of the text indicating a relevance of the comment to each segment, comprises applying natural language processing techniques to identify semantically related words and phrases appearing in the comment and the segment, respectively.

5. A method in accordance with claim 2 , wherein topics and concepts in the text are represented in the form of a mind map.

6. A method in accordance with claim 1 , further comprising converting, by the computer, informative comments selected for presentation to a user's preferred language in text, audio, or video form.

7. A computer program product for automatically augmenting online content with contextually relevant comments, the computer program product comprising:

one or more computer-readable non-transitory storage media and program instructions stored on the one or more non-transitory computer-readable storage media, the program instructions comprising:

program instructions to receive from a network server that hosts an online content provider, online content from the online content provider;

program instructions to receive from the online content provider, user-entered comments associated with the online content;

program instructions to classify the comments a trained machine learning model as either chit-chat if the comment corresponds to training data characterized as meaningless, unhelpful, or unimportant, or informative if the comment corresponds to training data characterized as containing valuable information;

for each comment classified as informative, program instructions to:

determine a portion of the online content to which the comment is most relevant, wherein the program instructions further comprise program instructions to:

convert the online content data to text;

assign scores to segments of the text indicating a relevance of the comment to each segment, and identify a segment with the highest score; and

associate the comment with a position in the online content that corresponds to the determined portion of the online content;

program instruction to select a subset of the informative comments for presentation based on a predefined criteria, wherein the predefined criteria comprises a selection based on: one or more of informative comments having a score exceeding a predefined threshold, recentness, randomness, a known or inferred profile of a user, a probabilistic basis according to a predefined distribution, or in real time, concurrently with user interaction with the online content;

program instructions to convert the selected subset of the informative commands from text to a format suitable for presentation with the online content data; and

program instructions to annotate the online content with the selected subset of informative comments at the positions in the online content associated with respective comments in the subset of informative comments.

8. A computer program product in accordance with claim 7 , wherein the online content is audio or video data, further comprising:

program instructions to convert speech in the online content to text; and wherein program instructions to determine comprise program instructions to determine a portion of the text to which the comment is most relevant.

9. A computer program product in accordance with claim 8 , wherein program instructions to determine a portion of the text to which the comment is most relevant comprise program instructions to assign scores to segments of the text indicating a relevance of the comment to each segment, and program instructions to identify a segment with the highest score.

10. A computer program product in accordance with claim 9 , wherein program instructions to assign scores to segments of the text indicating a relevance of the comment to each segment comprise program instructions to apply natural language processing techniques to identify semantically related words and phrases appearing in the comment and the segment, respectively.

11. A computer program product in accordance with claim 8 , wherein topics and concepts in the text are represented in the form of a mind map.

12. A computer program product in accordance with claim 7 , further comprising program instructions to convert informative comments selected for presentation to a user's preferred language in text, audio, or video form.

13. A computer system for automatically augmenting online content with contextually relevant comments, the computer system comprising:

one or more computer processors, one or more computer-readable storage media devices, and program instructions stored on one or more of the computer-readable storage media devices for execution by at least one of the one or more processors, the program instructions comprising:

program instructions to receive from a network server that hosts an online content provider, online content from the online content provider;

program instructions to receive from the online content provider, user-entered comments associated with the online content;

program instructions to classify the comments a trained machine learning model as either chit-chat if the comment corresponds to training data characterized as meaningless, unhelpful, or unimportant, or informative if the comment corresponds to training data characterized as containing valuable information;

for each comment classified as informative, program instructions to:

determine a portion of the online content to which the comment is most relevant, wherein the program instructions further comprise program instructions to:

convert the online content data to text;

assign, by the computer, scores to segments of the text indicating a relevance of the comment to each segment, and identifying a segment with the highest score; and

associate the comment with a position in the online content that corresponds to the determined portion of the online content;

program instruction to select a subset of the informative comments for presentation based on a predefined criteria, wherein the predefined criteria comprises a selection based on one or more of: informative comments having a score exceeding a predefined threshold, recentness, randomness, a known or inferred profile of a user, a probabilistic basis according to a predefined distribution, or in real time, concurrently with user interaction with the online content;

program instructions to convert the selected subset of the informative commands from text to a format suitable for presentation with the online content data; and

program instructions to annotate the online content with the selected subset of informative comments at the positions in the online content associated with respective comments in the subset of informative comments.

14. A computer program product in accordance with claim 13 , wherein the online content is audio or video data, further comprising:

program instructions to convert speech in the online content to text; and wherein program instructions to determine comprise program instructions to determine a portion of the text to which the comment is most relevant.

15. A computer program product in accordance with claim 14 , wherein program instructions to determine a portion of the text to which the comment is most relevant comprise program instructions to assign scores to segments of the text indicating a relevance of the comment to each segment, and program instructions to identify a segment with the highest score.

16. A computer program product in accordance with claim 15 , wherein program instructions to assign scores to segments of the text indicating a relevance of the comment to each segment comprise program instructions to apply natural language processing techniques to identify semantically related words and phrases appearing in the comment and the segment, respectively.

17. A computer program product in accordance with claim 13 , further comprising program instructions to convert informative comments selected for presentation to a user's preferred language in text, audio, or video form.