Abstract. The open nature of the World Wide Web makes evaluating webpage credibility challenging for users. In this paper, we aim to automatically assess web credibility by investigating various characteristics of webpages. Specifically, we first identify features from textual content, link structure, webpages design, as well as their social popularity learned from popular social media sites (e.g., Facebook, Twitter). A set of statistical analyses methods are applied to select the most informative features, which are then used to infer webpages credibility by employing supervised learning algorithms. Real dataset-based experiments under two application settings show that we attain an accuracy of 75% for classification, and an improvement of 53% for the mean absolute error (MAE), with respect to the random baseline approach, for regression.

We explore active learning (AL) for improving the accuracy of new domains in a natural language understanding (NLU) system. We propose an algorithm called Majority-CRF that uses an ensemble of classification models to guide the selection of relevant utterances, as well as a sequence labeling model to help prioritize informative examples. Experiments with three domains show that Majority-CRF achieves 6.6%-9% relative error rate reduction compared to random sampling with the same annotation budget, and statistically significant improvements compared to other AL approaches. Additionally, case studies with human-in-the-loop AL on six new domains show 4.6%-9% improvement on an existing NLU system.

No abstract

No abstract

Large transformer models, such as BERT, achieve state-of-the-art results in machine reading comprehension (MRC) for open-domain question answering (QA). However, transformers have a high computational cost for inference which makes them hard to apply to online QA systems for applications like voice assistants. To reduce computational cost and latency, we propose decoupling the transformer MRC model into input-component and crosscomponent. The decoupling allows for part of the representation computation to be performed offline and cached for online use. To retain the decoupled transformer accuracy, we devised a knowledge distillation objective from a standard transformer model. Moreover, we introduce learned representation compression layers which help reduce by four times the storage requirement for the cache. In experiments on the SQUAD 2.0 dataset, a decoupled transformer reduces the computational cost and latency of open-domain MRC by 30-40% with only 1.2 points worse F1-score compared to a standard transformer.