Abstract: Technologies pertaining to slot filling are described herein. A deep neural network, a recurrent neural network, and/or a spatio-temporally deep neural network are configured to assign labels to words in a word sequence set forth in natural language. At least one label is a semantic label that is assigned to at least one word in the word sequence.

Abstract: An understanding model is trained to account for human perception of the perceived relative importance of different tagged items (e.g. slot/intent/domain). Instead of treating each tagged item as equally important, human perception is used to adjust the training of the understanding model by associating a perceived weight with each of the different predicted items. The relative perceptual importance of the different items may be modeled using different methods (e.g. as a simple weight vector, a model trained using features (lexical, knowledge, slot type, . . . ), and the like). The perceptual weight vector and/or or model are incorporated into the understanding model training process where items that are perceptually more important are weighted more heavily as compared to the items that are determined by human perception as less important.

Abstract: Non-verbalized tokens, such as punctuation, are automatically predicted and inserted into a transcription of speech in which the tokens were not explicitly verbalized. Token prediction may be integrated with speech decoding, rather than performed as a post-process to speech decoding.

Abstract: A scalable statistical language understanding (SLU) system uses a fixed number of understanding models that scale across domains and intents (i.e. single vs. multiple intents per utterance). For each domain added to the SLU system, the fixed number of existing models is updated to reflect the newly added domain. Information that is already included in the existing models and the corresponding training data may be re-used. The fixed models may include a domain detector model, an intent action detector model, an intent object detector model and a slot/entity tagging model. A domain detector identifies different domains identified within an utterance. All/portion of the detected domains are used to determine associated intent actions. For each determined intent action, one or more intent objects are identified. Slot/entity tagging is performed using the determined domains, intent actions, and intent object detector.

Abstract: An orphan detector. The orphan detector processes out-of-domain utterances from a targeted language understanding dialog system to determine whether the out-of-domain utterance expresses a specific intent to have the targeted language understanding dialog system to take a certain action where fallback processing, such as performing a generic web search, is unlikely to be satisfied by web searches. Such utterances are referred to as orphans because they are not appropriately handled by any of the task domains or fallback processing. The orphan detector distinguishes orphans from web search queries and other out-of-domain utterances by focusing primarily on the structure of the utterance rather than the content. Orphans detected by the orphan detector may be used both online and offline to improve user experiences with targeted language understanding dialog systems. The orphan detector may also be used to mine structurally similar queries or sentences from the web search engine query logs.

Abstract: Non-verbalized tokens, such as punctuation, are automatically predicted and inserted into a transcription of speech in which the tokens were not explicitly verbalized. Token prediction may be integrated with speech decoding, rather than performed as a post-process to speech decoding.

Abstract: Technologies pertaining to slot filling are described herein. A deep neural network, a recurrent neural network, and/or a spatio-temporally deep neural network are configured to assign labels to words in a word sequence set forth in natural language. At least one label is a semantic label that is assigned to at least one word in the word sequence.

Abstract: Non-verbalized tokens, such as punctuation, are automatically predicted and inserted into a transcription of speech in which the tokens were not explicitly verbalized. Token prediction may be integrated with speech decoding, rather than performed as a post-process to speech decoding.

Abstract: An understanding model is trained to account for human perception of the perceived relative importance of different tagged items (e.g. slot/intent/domain). Instead of treating each tagged item as equally important, human perception is used to adjust the training of the understanding model by associating a perceived weight with each of the different predicted items. The relative perceptual importance of the different items may be modeled using different methods (e.g. as a simple weight vector, a model trained using features (lexical, knowledge, slot type, . . . ), and the like). The perceptual weight vector and/or or model are incorporated into the understanding model training process where items that are perceptually more important are weighted more heavily as compared to the items that are determined by human perception as less important.

Abstract: A scalable statistical language understanding (SLU) system uses a fixed number of understanding models that scale across domains and intents (i.e. single vs. multiple intents per utterance). For each domain added to the SLU system, the fixed number of existing models is updated to reflect the newly added domain. Information that is already included in the existing models and the corresponding training data may be re-used. The fixed models may include a domain detector model, an intent action detector model, an intent object detector model and a slot/entity tagging model. A domain detector identifies different domains identified within an utterance. All/portion of the detected domains are used to determine associated intent actions. For each determined intent action, one or more intent objects are identified. Slot/entity tagging is performed using the determined domains, intent actions, and intent object detector.

Abstract: A method and system for automatic gain control of a speech signal in a communication system are disclosed. The gain of the speech signal can be controlled, based on a calculated gain value. This gain value is calculated on the basis of energy calculation and speech activity identification in the speech signal which is done by means of the encoder. Encoding the gain controlled speech signal for transmission follows the step of gain control.

Abstract: Non-verbalized tokens, such as punctuation, are automatically predicted and inserted into a transcription of speech in which the tokens were not explicitly verbalized. Token prediction may be integrated with speech decoding, rather than performed as a post-process to speech decoding.

Abstract: A method and system for automatic gain control of a speech signal in a communication system are disclosed. The gain of the speech signal can be controlled, based on a calculated gain value. This gain value is calculated on the basis of energy calculation and speech activity identification in the speech signal which is done by means of the encoder. Encoding the gain controlled speech signal for transmission follows the step of gain control.