Mengqian Wang

Mengqian Wang, Ilya Valmianski, Xavier Amatriain et al.

Medical conversations between patients and medical professionals have implicit functional sections, such as "history taking", "summarization", "education", and "care plan." In this work, we are interested in learning to automatically extract these sections. A direct approach would require collecting large amounts of expert annotations for this task, which is inherently costly due to the contextual inter-and-intra variability between these sections. This paper presents an approach that tackles the problem of learning to classify medical dialogue into functional sections without requiring a large number of annotations. Our approach combines pseudo-labeling and human-in-the-loop. First, we bootstrap using weak supervision with pseudo-labeling to generate dialogue turn-level pseudo-labels and train a transformer-based model, which is then applied to individual sentences to create noisy sentence-level labels. Second, we iteratively refine sentence-level labels using a cluster-based human-in-the-loop approach. Each iteration requires only a few dozen annotator decisions. We evaluate the results on an expert-annotated dataset of 100 dialogues and find that while our models start with 69.5% accuracy, we can iteratively improve it to 82.5%. The code used to perform all experiments described in this paper can be found here: https://github.com/curai/curai-research/tree/main/functional-sections.

Xuhang Chen, Bo Lv, Mengqian Wang et al.

Customer churn prediction in the telecommunications sector represents a critical business intelligence task that has evolved from subjective human assessment to sophisticated algorithmic approaches. In this work, we present a comprehensive framework for telecommunications churn prediction leveraging deep neural networks. Through systematic problem formulation, rigorous dataset analysis, and careful feature engineering, we develop a model that captures complex patterns in customer behavior indicative of potential churn. We conduct extensive empirical evaluations across multiple performance metrics, demonstrating that our proposed neural architecture achieves significant improvements over existing baseline methods. Our approach not only advances the state-of-the-art in churn prediction accuracy but also provides interpretable insights into the key factors driving customer attrition in telecommunications services.

Michael Segundo Ortiz, Sam Bubnovich, Mengqian Wang et al.

The largest collection of medical evidence in the world is PubMed. However, the significant barrier in accessing and extracting information is information organization. A factor that contributes towards this barrier is managing medical controlled vocabularies that allow us to systematically and consistently organize, index, and search biomedical literature. Additionally, from users' perspective, to ultimately improve access, visualization is likely to play a powerful role. There is a strong link between information organization and information visualization, as many powerful visualizations depend on clustering methods. To improve visualization, therefore, one has to develop concrete and scalable measures for vocabularies used in indexing and their impact on document clustering. The focus of this study is on the development and evaluation of clustering methods. The paper concludes with demonstration of downstream network visualizations and their impact on discovering potentially valuable and latent genetic and molecular associations.