Publications

Recommending a diversity of product types (PTs) is important for a good shopping experience when customers are looking for products around their high-level shopping interests (SIs) such as hiking. However, the SI-PT connection is typically absent in e-commerce product catalogs and expensive to construct manually due to the volume of potential SIs, which prevents us from establishing a recommender with easily

Graph Neural Networks (GNNs) have gained popularity in various fields, such as recommendation systems, social network analysis and fraud detection. However, despite their effectiveness, the topological nature of GNNs makes it challenging for users to understand the model predictions. To address this challenge, we built a user-friendly UI to visualize the most important relationships for both homogeneous

NoSQL cloud database services are popular for their simple key-value operations, high availability, high scalability, and predictable performance. These characteristics are generally considered to be at odds with support for transactions that permit atomic and serializable updates to partitioned data. This paper explains how transactions were added to Amazon DynamoDB using a timestamp ordering protocol

The ability to jointly learn from multiple modalities, such as text, audio, and visual data, is a defining feature of intelligent systems. While there have been promising advances in designing neural networks to harness multimodal data, the enormous success of data augmentation currently remains limited to single-modality tasks like image classification. Indeed, it is particularly difficult to augment each

Recently, neural models have been leveraged to significantly improve the performance of information extraction from semi-structured websites. However, a barrier for continued progress is the small number of datasets large enough to train these models. In this work, we introduce the PLAtE (Pages of Lists Attribute Extraction) benchmark dataset as a challenging new web extraction task. PLAtE focuses on shopping

Recent advancements in generative models have led to significant improvements in the quality of generated images, making them virtually indistinguishable from real ones. However, using AI generated images for training robust computer vision models for real-world applications, especially object detection in road scene perception, is still a challenge. AI generated images usually lack the required diversity

Neural network implementations have predominantly been a black box lacking both in interpretability and estimation of uncertainty. In this study, we propose a novel causal attribution methodology for mixture density networks wherein we outline a framework to compute the causal effect of each feature on the target variable along with the associated uncertainty in the attribution. Our approach allows for

Information Extraction (IE) from semi-structured web-pages is a long studied problem. Training a model for this extraction task requires a large number of human-labeled samples. Prior works have proposed transferable models to improve the label-efficiency of this training process. Extraction performance of transferable models however, depends on the size of their fine-tuning corpus. This holds true for

Low-count time series describe sparse or intermittent events, which are prevalent in large-scale online platforms that capture and monitor diverse data types. Several distinct challenges surface when modelling low-count time series, particularly low signal-to-noise ratios (when anomaly signatures are provably undetectable), and non-uniform performance (when average metrics are not representative of local

We propose a methodology for information aggregation from the various transformer layer outputs of a generic speech Encoder (e.g. WavLM, HuBERT) for the downstream task of Speech Emotion Recognition (SER). The proposed methodology significantly reduces the dependency of model predictions on linguistic content, while leading to competitive performance without requiring costly Encoder re-training. The proposed

The issue of enhancing the robustness of Named Entity Recognition (NER) models against adversarial attacks has recently gained significant attention (Simoncini and Spanakis, 2021; Lin et al., 2021). The existing techniques for robustifying NER models rely on exhaustive perturbation of the input training data to generate adversarial examples, often resulting in adversarial examples that are not semantically

Personalization plays a critical role in helping customers discover the products and contents they prefer for e-commerce stores. Personalized recommendations differ in contents, target customers, and UI. However, they require a common core capability - the ability to deeply understand customers’ preferences and shopping intents. In this paper, we introduce the MCM (Multi-task pre-trained Customer Model)

Audio privacy has been undertaken using adversarial task training or adversarial models based on GANs, where the models also suppress scoring of other attributes (e.g., emotion, etc.), but embeddings still retain enough information to bypass speaker privacy. We use methods for feature importance from the explainability literature to modify embeddings from adversarial task training, providing a simple and

Speech super-resolution is the process of estimating the missing frequency content of a speech signal from its existing band-limited frequency content. The loss of frequency components is a common occurrence that can be because of a low sampling rate, low-quality microphones, or various transmission factors, and it is an increasingly common problem as bandwidth for high-quality communications is generally

We present a new task setting for attribute mining on e-commerce products, serving as a practical solution to extract open-world attributes without extensive human intervention. Our supervision comes from a high-quality seed attribute set bootstrapped from existing resources, and we aim to expand the attribute vocabulary of existing seed types, and also to discover any new attribute types automatically.