Experience

• Arizona State University

  Research Assistant

  Sentiment Analysis for Short Texts in Social Media (in collaboration with Yahoo! Labs). We propose machine learning models to address two challenges that the texts are (1) noisy, and (2) sparse & short, (WSDM'13) by taking advantage of that the data instances that are (3) networked, and (4) with abundant contextual information (WWW'13).
  
  Active Learning for Big Data in Social Media (SDM'13).
  
  User Behavior Analysis in Online E-commerce. (WSDM'13b)
  
  Analysis of Palin's Email Network. We Contribute an investigation of social status and role analysis on Sarah Palin's email corpus from three different perspectives: individual, group, and temporal. (WWW'12)
  
  Enhancing Accessibility of Microblogging Messages. We Propose to perform microblogging messages clustering and labeling by utilizing tree kernel and semantic knowledge from Wikipedia and WordNet. (CIKM'11)

• National University of Singapore

  Research Assistant

  Aggregated Contextual Search (Metasearch). I participate in designing and developing the query recommendation module, the query transformation, and the clustering module of the aggregated search system. Short Texts Clustering. We propose a novel framework to improve the performance of short texts clustering by exploiting the internal semantics from original text and external concepts from Wikipedia and WordNet.

• Beihang University

  Research Assistant

  Xia worked at Beihang University as a Research Assistant

• Texas A&M University

  Assistant Professor

  Automated Machine Learning (AutoML): We have Pre-released Auto-Keras system (over 5,000 stars and 800 forks on Github) on automated machine learning.
  
  Explainable Artificial Intelligence (XAI): We develop algorithms and systems to understand how Machine Learning systems work and how a decision is made through complicated black-box models.
  
  Network Analytics: We develop effective and efficient algorithms to analyze large-scale, dynamic, heterogeneous, and sparse network data such as social networks, patient networks, purchasing networks in e-commerce.
  
  Recommender Systems: The NCF package has become the official recommender system in TensorFlow.
  
  Anomaly Detection: We have been working with Apple, Ingersoll Rand, UnitedHealthcare and JP Morgan for anomaly detection on different industrial applications.

• Arizona State University and Phoenix VA Health Care System

  Postdoctoral Researcher

  Xia worked at Arizona State University and Phoenix VA Health Care System as a Postdoctoral Researcher

• Microsoft

  Research Intern

  Entity Extraction and Relation Modeling. We work on a project to extract entities, including people, location, date, etc., and the relations between each other. The goal is to build the knowledge base to help Bing power the next generation of entity centric experiences and recommendation system.