Advertisement placement and engagement supported video indexing, search and delivery platforms, methods, systems and apparatuses
Advertisements are delivered with videos, the advertisements being cued to begin based on an analysis of metadata associated with the videos such as moments identified from user-submitted comments relative to the video. Credit earned by users based on their viewing and engagement activities may be gifted to other users or applied towards baseline fees assessed in connection with videos delivered to the user.
1. A method for delivering video and advertisements comprising:
selectively ingesting metadata including one or more comments corresponding to one or more videos;
analyzing the ingested video metadata, by a metadata analysis component executed by a processor, to identify moments in the corresponding video by parsing comments relating to the video, identifying a moment if a comment includes a reference to a time in the video and storing the time and a text of the comment in a record of the moment;
evaluating videos, by an evaluation component executed by a processor, based on a comparison of their corresponding moments to advertisement parameters to identify one or more videos appropriate for an advertisement;
analyzing the moments corresponding to the one or more videos identified as being appropriate for an advertisement, by a moment analysis component executed by a processor, to identify is one or more advertisement cues;
delivering the one or more videos identified as being appropriate for an advertisement along with an advertisement, the advertisement being cued to begin at a time relative to the one or more videos based on the one or more identified advertisement cues;
wherein an advertisement cue for a particular video is identified at the most moment-dense portion of the video, and
wherein the most moment-dense portion of the video is identified by segmenting the video into discrete time segments and counting the number of moments falling within each time segment based on the moments' corresponding times, identifying the time segment with the highest number of moments as the most moment-dense.
2. The method of claim 1 , further comprising: setting back the advertisement beginning time by a predetermined setback from the one or more identified advertisement cues.
3. The method of claim 1 , wherein two or more advertisement cues for a particular video are identified at the most moment-dense portions of the video.
4. The method of claim 1 , wherein: before the moments are analyzed to identify advertisement cues, the moments are filtered based on the text of the moments to identify filtered moments that meet the advertisement parameters and advertisement cues are identified based on an analysis of the filtered moments.
5. The method of claim 4 , wherein the filtering includes identifying a sentiment of the text of the moments and filtering the moments based on a sentiment identified by the advertisement parameters.
6. A method for delivering video and advertisements comprising:
selectively ingesting metadata including one or more comments corresponding to one or more videos;
analyzing the ingested video metadata, by a metadata analysis component executed by a processor, to identify moments in the corresponding video by parsing comments relating to the video, identifying a moment if a comment includes a reference to a time in the video and storing the time and a text of the comment in a record of the moment;
analyzing the moments corresponding to the one or more videos, by a metadata analysis component executed by a processor, to identify one or more advertisement cues; and
delivering the one or more videos along with an advertisement, the advertisement being cued to begin at a time relative to the one or more videos based on the one or more identified advertisement cue;
wherein an advertisement cue for a particular video is identified at the most moment-dense portion of the video, and
wherein the most moment-dense portion of the video is identified by segmenting the video into discrete time segments and counting the number of moments falling within each time segment based on the moments' corresponding times, identifying the time segment with the highest number of moments as the most moment-dense.
7. The method of claim 6 , further comprising: setting back the advertisement beginning time by a predetermined setback from the one or more identified advertisement cues.
8. The method of claim 6 , wherein: before the moments are analyzed to identify advertisement cues, the moments are filtered based on the text of the moments to identify filtered moments that meet the advertisement parameters and advertisement cues are identified based on an analysis of the filtered moments.
9. The method of claim 8 , wherein the filtering includes identifying a sentiment of the text of the moments and filtering the moments based on a sentiment identified by the advertisement parameters.