Abstract: At an artificial intelligence system, training iterations of a first machine learning model are implemented. In a particular iteration, a group of data items are selected from an item collection using active learning, and respective labels selected from a set of tags are obtained for at least some of the items of the group. Using feature processing elements of a different machine learning model, a respective feature set corresponding to individual labeled items is generated in the iteration, and the feature sets are included in a training set used to train the first machine learning model. A trained version of the first machine learning model is stored after a training completion criterion is met.

Abstract: At an artificial intelligence system, training iterations of a first machine learning model are implemented. In a particular iteration, a group of data items are selected from an item collection using active learning, and respective labels selected from a set of tags are obtained for at least some of the items of the group. Using feature processing elements of a different machine learning model, a respective feature set corresponding to individual labeled items is generated in the iteration, and the feature sets are included in a training set used to train the first machine learning model. A trained version of the first machine learning model is stored after a training completion criterion is met.

Abstract: Techniques for generating a recommendation for content based on user-device dialogue are described. In an example, a computer system receives a request for audio output at a user device. The computer system determines a response for the request. The computer system determines a topic and value associated with the request and the response. The computer system also determines a recommendation for content based on the request and the response. The computer system generates prompt data including a question of whether to initiate an action associated with the recommendation. The computer system sends the prompt data and the response data to the user device or a different user device.

Abstract: Techniques for generating a recommendation for content based on user-device dialogue are described. In an example, a computer system receives a request for audio output at a user device. The computer system determines a response for the request. The computer system determines a topic and value associated with the request and the response. The computer system also determines a recommendation for content based on the request and the response. The computer system generates prompt data including a question of whether to initiate an action associated with the recommendation. The computer system sends the prompt data and the response data to the user device or a different user device.

Abstract: Techniques for identifying and correcting synchronization errors for a media file are described herein. A first file that includes a first set of words comprising lyrics of a media file may be maintained. One or more portions of the media file that represent vocal audio may be separated from other portions of the media file that represent instrumental audio by a computer system. A second file may be generated based at least in part on using an automated speech recognition on the separated one or more portions of the media file. The second file my include time stamps for a second set of words comprising the lyrics in the separated one or more portions of the media file. The first file may be modified with an offset time value that is determined by aligning the first set of words with the second set of words.

Abstract: A technique for determining annotation items associated with video information is described. During this annotation technique, a content item that includes audio information and the video information is received. For example, a file may be downloaded from a uniform resource locator. Then, the audio information is extracted from the content item, and the audio information is analyzed to determine features or descriptors that characterize the audio information. Note that the features may be determined solely by analyzing the audio information or may be determined by subsequent further analysis of at least some of the video information based on the analysis of the audio information (i.e., sequential or cascaded analysis). Next, annotation items or tags associated with the video information are determined based on the features.

Abstract: A technique for determining annotation items associated with video information is described. During this annotation technique, a content item that includes audio information and the video information is received. For example, a file may be downloaded from a uniform resource locator. Then, the audio information is extracted from the content item, and the audio information is analyzed to determine features or descriptors that characterize the audio information. Note that the features may be determined solely by analyzing the audio information or may be determined by subsequent further analysis of at least some of the video information based on the analysis of the audio information (i.e., sequential or cascaded analysis). Next, annotation items or tags associated with the video information are determined based on the features.

Abstract: A technique for determining annotation items associated with video information is described. During this annotation technique, a content item that includes audio information and the video information is received. For example, a file may be downloaded from a uniform resource locator. Then, the audio information is extracted from the content item, and the audio information is analyzed to determine features or descriptors that characterize the audio information. Note that the features may be determined solely by analyzing the audio information or may be determined by subsequent further analysis of at least some of the video information based on the analysis of the audio information (i.e., sequential or cascaded analysis). Next, annotation items or tags associated with the video information are determined based on the features.