Abstract: In an example embodiment, career path information is extracted from user data, the career path information including two or more career path features for each job held by the user, as identified in the career path information. The career path information is concatenated into a single concatenation. The single concatenation is fed into an embedding model trained using a first machine learning algorithm to output one or more embeddings in an n-dimensional space for the single concatenation. These embeddings are then used as input to a model trained by a second machine learning algorithm.

Abstract: The disclosed embodiments relate to the design of a system that uses an accelerometer in a mobile device to detect hotwords, which activate a voice interface to recognize subsequent voice input. During operation, the system gathers samples comprising readings from the accelerometer in the mobile device. Next, the system calculates features from the gathered samples. The system then classifies the calculated features using a classifier, which has been trained to detect hotwords based on calculated features. Finally, if the classifier detects one or more hotwords, the system causes the mobile device to launch a voice-control system that recognizes subsequent voice input received from an audio microphone.

Abstract: The disclosed embodiments relate to the design of a system that uses an accelerometer in a mobile device to detect hotwords, which activate a voice interface to recognize subsequent voice input. During operation, the system gathers samples comprising readings from the accelerometer in the mobile device. Next, the system calculates features from the gathered samples. The system then classifies the calculated features using a classifier, which has been trained to detect hotwords based on calculated features. Finally, if the classifier detects one or more hotwords, the system causes the mobile device to launch a voice-control system that recognizes subsequent voice input received from an audio microphone.