Abstract: Embodiments described herein provide an attention-based tree encoding mechanism. Specifically, the attention layer receives as input the pre-parsed constituency tree of a sentence and the lower-layer representations of all nodes. The attention layer then performs upward accumulation to encode the tree structure from leaves to the root in a bottom-up fashion. Afterwards, weighted aggregation is used to compute the final representations of non-terminal nodes.

Abstract: A method provides learning with noisy labels. The method includes generating a first network of a machine learning model with a first set of parameter initial values, and generating a second network of the machine learning model with a second set of parameter initial values. First clean probabilities for samples in a training dataset are generated using the second network. A first labeled dataset and a first unlabeled dataset are generated from the training dataset based on the first clean probabilities. The first network is trained based on the first labeled dataset and first unlabeled dataset to update parameters of the first network.

Abstract: Embodiments described in this disclosure illustrate the use of self-/semi supervised approaches for label-efficient DST in task-oriented dialogue systems. Conversational behavior is modeled by next response generation and turn utterance generation tasks. Prediction consistency is strengthened by augmenting data with stochastic word dropout and label guessing. Experimental results show that by exploiting self-supervision the joint goal accuracy can be boosted with limited labeled data.

Abstract: Embodiments described herein provide a dynamic topic tracking mechanism that tracks how the conversation topics change from one utterance to another and use the tracking information to rank candidate responses. A pre-trained language model may be used for response selection in the multi-party conversations, which consists of two steps: (1) a topic-based pre-training to embed topic information into the language model with self-supervised learning, and (2) a multi-task learning on the pretrained model by jointly training response selection and dynamic topic prediction and disentanglement tasks.

Abstract: A system and method for translating questions into database queries are provided. A text to database query system receives a natural language question and a structure in a database. Question tokens are generated from the question and query tokens are generated from the structure in the database. The question tokens and query tokens are concatenated into a sentence and a sentence token is added to the sentence. A BERT network generates question hidden states for the question tokens, query hidden states for the query tokens, and a classifier hidden state for the sentence token. A translatability predictor network determines if the question is translatable or untranslatable. A decoder converts a translatable question into an executable query. A confusion span predictor network identifies a confusion span in the untranslatable question that causes the question to be untranslatable. An auto-correction module to auto-correct the tokens in the confusion span.

Abstract: A method for dialog state tracking includes decoding, by a fertility decoder, encoded dialog information associated with a dialog to generate fertilities for generating dialog states of the dialog. Each dialog state includes one or more domains. Each domain includes one or more slots. Each slot includes one or more slot tokens. The method further includes generating an input sequence to a state decoder based on the fertilities. A total number of each slot token in the input sequence is based on a corresponding fertility. The method further includes encoding, by a state encoder, the input sequence to the state decoder, and decoding, by the state decoder, the encoded input sequence to generate a complete sequence of the dialog states.

Abstract: A visual dialogue model receives image input and text input that includes a dialogue history between the model and a current utterance by a human user. The model generates a unified contextualized representation using a transformer encoder network, in which the unified contextualized representation includes a token level encoding of the image input and text input. The model generates an encoded visual dialogue input from the unified contextualized representation using visual dialogue encoding layers. The encoded visual dialogue input includes a position level encoding and a segment type encoding. The model generates an answer prediction from the encoded visual dialogue input using a first self-attention mask associated with discriminative settings or a second self-attention mask associated with generative settings. Dense annotation fine tuning may be performed to increase accuracy of the answer prediction. The model provides the answer prediction as a response to the current utterance of the human user.

Abstract: Embodiments described herein provide visual-and-language (V+L) systems and methods for learning vision and language representations. Specifically, a method may comprise receiving a training dataset comprising a plurality of image samples and a plurality of text samples; encoding the plurality of image samples into a plurality of encoded image samples and the plurality of text samples into a plurality of encoded text samples; computing a first loss objective based on the plurality of encoded image samples and the plurality of encoded text samples; encoding a first subset of the plurality of encoded image samples and a second subset of the plurality of encoded text samples into a plurality of encoded image-text samples; computing a second loss objective based on the plurality of encoded image-text samples; and updating the V+L model based at least in part on the first loss objective and the second loss objective.

Abstract: Embodiments are directed to pre-training a transformer model using more parameters for sophisticated patterns (PSP++). The transformer model is divided into a held-out model and a main model. A forward pass and a backward pass are performed on the held-out model, where the forward pass determines self-attention hidden states of the held-out model and the backward pass determines loss of the held-out model. A forward pass on the main model is performed to determine a self-attention hidden states of the main model. The self-attention hidden states of the main model are concatenated with the self-attention hidden states of the held-out model. A backward pass is performed on the main model to determine a loss of the main model. The parameters of the held-out model are updated to reflect the loss of the held-out model and parameters of the main model are updated to reflect the loss of the main model.

Abstract: Embodiments described herein provide a causality-based anomaly detection mechanism that formulates multivariate time series as instances that do not follow the regular causal mechanism. Specifically, the causality-based anomaly detection mechanism leverages the causal structure discovered from data so that the joint distribution of multivariate time series is factorized into simpler modules where each module corresponds to a local causal mechanism, reflected by the corresponding conditional distribution. Those local mechanisms are modular or autonomous and can then be handled separately. In light of this modularity property, the anomaly detection problem then naturally decomposed into a series of low-dimensional anomaly detection problems. Each sub-problem is concerned with a local mechanism.

Abstract: Embodiments described herein a code generation and understanding model that builds on a Transformer-based encoder-decoder framework. The code generation and understanding model is configured to derive generic representations for programming language (PL) and natural language (NL) in code domain via pre-training on unlabeled code corpus, and then to benefit many code-related downstream tasks with fine-tuning. Apart from the denoising sequence-to-sequence objectives widely adopted for pre-training on natural language, identifier tagging and prediction pre-training objective is adopted to enable the model to better leverage the crucial token type information from PL, which specifically are the identifiers assigned by developers.

Abstract: Embodiments described herein provides a contrastive learning framework that leverages hard negative examples, that are mined globally from the entire training corpus for a given query to improve the quality of code and natural language representations. Specifically, similar examples from the training corpus are extracted and used as hard negatives in an online manner during training while keeping the minibatch construction random.

Abstract: Some embodiments of the current disclosure disclose methods and systems for analyzing root causes of an incident disrupting information technology services such as cloud services. In some embodiments, a set of problem review board (PRB) documents including information about said incidents may be parsed using a natural language processing (NLP) neural model to extract structured PRB data from the unstructured investigative information contained in the PRB documents. The structured PRB data may include symptoms of the incident, root causes of the incident, resolutions of the incidents, etc., and a causal knowledge graph causally relating the symptoms, root causes, resolutions of the incidents may be generated.

Abstract: A system and method for generating a response in a video grounded dialogue are provided. A video-grounded dialogue neural network language model receives video input and text input. The text input includes a dialogue history between the model and a human user and a current utterance by the user. Encoded video input is generated using video encoding layers. Encoded text input is generated using text encoding layers. The encoded video input and the encoded text input are concatenated in to a single input sequence. A generative pre-trained transformer model generates the response to the current utterance from the singe input sequence.

Abstract: Embodiments described herein provide a contrastive learning mechanism with self-labeling refinement, which iteratively employs the network and data themselves to generate more accurate and informative soft labels for contrastive learning. Specifically, the contrastive learning framework includes a self-labeling refinery module to explicitly generate accurate labels, and a momentum mix-up module to increase similarity between a query and its positive, which in turn implicitly improves label accuracy.

Abstract: A multi-view contrastive relational learning framework is provided. In the multi-view contrastive relational learning framework, contrastive learning is augmented with a multi-view learning signal. The auxiliary views guide an encoder of the underlying time series data's main view, by using an inter-sample similarity structure as a learning signal to learn representations which encode information from multiple views.

Abstract: Embodiments described in this disclosure illustrate the use of self-/semi supervised approaches for label-efficient DST in task-oriented dialogue systems. Conversational behavior is modeled by next response generation and turn utterance generation tasks. Prediction consistency is strengthened by augmenting data with stochastic word dropout and label guessing. Experimental results show that by exploiting self-supervision the joint goal accuracy can be boosted with limited labeled data.

Abstract: A system uses a neural network based model to perform scene text recognition. The system achieves high accuracy of prediction of text from scenes based on a neural network architecture that uses double attention mechanism. The neural network based model includes a convolutional neural network component that outputs a set of visual features and an attention extractor neural network component that determines attention scores based on the visual features. The visual features and the attention scores are combined to generate mixed features that are provided as input to a character recognizer component that determines a second attention score and recognizes the characters based on the second attention score. The system trains the neural network based model by adjusting the neural network parameters to minimize a multi-class gradient harmonizing mechanism (GHM) loss. The multi-class GHM loss varies based on a level of difficulty of the sample.

Abstract: Embodiments described herein provide an approach (referred to as “Co-training” mechanism throughout this disclosure) that jointly learns two representations of the training data, their class probabilities and low-dimensional embeddings. Specifically, two representations of each image sample are generated: a class probability produced by the classification head and a low-dimensional embedding produced by the projection head. The classification head is trained using memory-smoothed pseudo-labels, where pseudo-labels are smoothed by aggregating information from nearby samples in the embedding space. The projection head is trained using contrastive learning on a pseudo-label graph, where samples with similar pseudo-labels are encouraged to have similar embeddings.

Abstract: Embodiments described herein provide systems and methods for learning representation from unlabeled videos. Specifically, a method may comprise generating a set of strongly-augmented samples and a set of weakly-augmented samples from the unlabeled video samples; generating a set of predictive logits by inputting the set of strongly-augmented samples into a student model and a first teacher model; generating a set of artificial labels by inputting the set of weakly-augmented samples to a second teacher model that operates in parallel to the first teacher model, wherein the second teacher model shares one or more model parameters with the first teacher model; computing a loss objective based on the set of predictive logits and the set of artificial labels; updating student model parameters based on the loss objective via backpropagation; and updating the shared parameters for the first teacher model and the second teacher model based on the updated student model parameters.