Systems and methods are described for modifying a phonetic search index based on a use frequency associated with phonetic representations of text terms included in metadata of a media item. A first phonetic representation of a text term of the metadata, pronounced as a word, may be generated. A second phonetic representation of the text term may be generated by concatenating a phonetic representation of each letter in the text term. A database may be queried to determine use frequencies of the first and second phonetic representations, one of which may be selected based on a comparison of the use frequencies. A phonetic search index may be modified by including an entry for the selected phonetic representation. A voice query related to the media item may be received, and a reply to the voice query may be generated for output by performing a lookup in the modified phonetic search index.

Methods and apparatus for improving speech recognition accuracy in media content searches are described. An advertisement for a media content item is analyzed to identify keywords that may describe the media content item. The identified keywords are associated with the media content item for use during a voice search to locate the media content item. A user may speak the one or more of the keywords as a search input and be provided with the media content item as a result of the search.

Methods and apparatus for improving speech recognition accuracy in media content searches are described. An advertisement for a media content item is analyzed to identify keywords that may describe the media content item. The identified keywords are associated with the media content item for use during a voice search to locate the media content item. A user may speak the one or more of the keywords as a search input and be provided with the media content item as a result of the search.

Methods, apparatus, systems and articles of manufacture are disclosed for a window-metric rule for return path data (RPD). In some examples, a media monitor is to determine a first count of a number of devices in a first household that reported RPD for a first day, and compare the first count to a current window-metric, the current window-metric associated with a window of time for the first household, the window of time associated with a content provider and having N number of days. In some examples, the media monitor is also to exclude the RPD from the first household for the first day when the first count does not satisfy the current window-metric, and generate a media exposure report using the RPD from the first household for the first day when the first count satisfies the current window-metric, independent of if the first household has heartbeat data available.

An electronic device obtains a plurality of collections of media content items, each collection of media content items being associated with text generated by one or more users of the media-providing service. The electronic device determines a coincidence metric for a first descriptor and a first media content item, the coincidence metric corresponding to a likelihood that the first descriptor appears in the text associated with a respective collection of media content items that includes the first media content item. Based on the coincidence metric, the electronic device generates a new collection of media content items for a first user. The new collection of media content items corresponds to the first descriptor and includes the first media content item.

Techniques for personalized gamification of media content. An engagement level of a consumer is identified based on prior gamification data. A difficulty level is identified using machine learning based on the engagement level. A content element personalized for the consumer is generated based on biographical information or viewing habits. A prompt requesting the consumer to find and access multimedia content scenes depicting the content element is generated. A multimedia content scene found by the consumer is analyzed to determine whether the multimedia content scene has an association with the content element and whether the consumer has accessed the multimedia content element scene.

Techniques for personalized gamification of media content. An engagement level of a consumer is identified based on prior gamification data. A difficulty level is identified using machine learning based on the engagement level. A content element personalized for the consumer is generated based on biographical information or viewing habits. A prompt requesting the consumer to find and access multimedia content scenes depicting the content element is generated. A multimedia content scene found by the consumer is analyzed to determine whether the multimedia content scene has an association with the content element and whether the consumer has accessed the multimedia content element scene.

A hardware media items scheduling and packaging system, which schedules and distributes channels to be viewed on a plurality of consumer devices, extracts contextual data from program-specific information associated with programming media content of a channel received from a distribution source device. A plurality of potential non-programming media items is determined for a plurality of users based on a match between a sentiment type of each of a plurality of non-programming media items and the extracted contextual data. Based on at least the extracted contextual data and the sentiment type of each of the plurality of potential non-programming media items, a plurality of candidate spots in the programming media content is determined. Based on at least a set of constraints and user estimation data associated with the plurality of users, a schedule of non-programming media item(s) is dynamically generated for at least one candidate spot in the programming media content.

A hardware media items scheduling and packaging system, which schedules and distributes channels to be viewed on a plurality of consumer devices, extracts contextual data from program-specific information associated with programming media content of a channel received from a distribution source device. A plurality of potential non-programming media items is determined for a plurality of users based on a match between a sentiment type of each of a plurality of non-programming media items and the extracted contextual data. Based on at least the extracted contextual data and the sentiment type of each of the plurality of potential non-programming media items, a plurality of candidate spots in the programming media content is determined. Based on at least a set of constraints and user estimation data associated with the plurality of users, a schedule of non-programming media item(s) is dynamically generated for at least one candidate spot in the programming media content.

According to one implementation, a system for performing graph-based media content evaluation includes a computing platform having a hardware processor, and a system memory storing a media content evaluation software code and a graph database. The hardware processor is configured to execute the media content evaluation software code to receive a query from a system user, and to identify one or more media content evaluation metrics corresponding to the query. In addition, the hardware processor is configured to execute the media content evaluation software code to search the graph database for a media content data relevant to the one or more media content evaluation metrics, and to retrieve the media content data from the graph database. The hardware processor is further configured to execute the media content evaluation software code to generate a report responsive to the query using the media content data.