Lifelong Machine Learning (Second Edition) Download for free, if your institution is on this list ----- Get the first edition. Morgan & Claypool Publishers, 207 pages, August, 2018. Order from Amazon.com

Sentiment Analysis: mining opinions, sentiments, and emotions How to build a Sentiment Analysis system, and much more Table of Contents, 381 pages Cambridge Univ. Press, June 2015. Order from Amazon.com

Sentiment Analysis and Opinion Mining A comprehensive introduction and survey Morgan & Claypool, May 2012, 167 pages Order from Amazon.com

Web Data Mining: exploring hyperlinks, contents, and usage data 2nd Edition, 2011, 622 Pages Order from Amazon.com, Order from Springer Get the eBook

Research Projects

• Opinion Mining, Sentiment Analysis and Opinion Spam Detection
  • Book: Sentiment Analysis and Opinion Mining: Introduction and survey
  • Emotional chatbot - enable chatbots or dialogue systems to converse with emotions
  • Reviews Reading Comprehension: Answer questions using reviews - When you plan to buy a product, do you have questions? Yes, 100%.
• Lifelong Machine Learning - Continual Learning
  • Continual Learning, Continuous Learning and Meta-Learning: Learn like humans - accumulating the prevously learned knowledge and adapt/transfer it to help future learning.
  • Second edition: Lifelong Machine Learning, Morgan & Claypool publishers, 207 pages, Aug 2018 (1st edition, November 2016),
  • Lifelong Machine Learning Tutorial, IJCAI-2015, July 25-31, 2015, Buenos Aires, Argentina
• Open World Machine Learning, Classification, or Recognition: one form of Lifelong Learning
  • Closed world assumption: classic machine learning assumes that test instances must be from training classes. Not true in many real-world environments.
  • Open-world learning: Identify instances of unseen classes during testing or application and incrementally learn the new classes.
• Continuous Knowledge Learning in Dialogues
  • To enable a dialgoue system (chatbot) to learn new knowledge continuously during conversations (after deployment), i.e., learning on the job.
• PU Learning - Learning from Positive and Unlabeled Examples (Set Expansion is also a form of PU learning)


• Earlier projects (inactive currently)
  • Detect Fake Reviews and Fake Reviewers - Detect Opinion Spam
    • Detecting fake or deceptive opinions on review sites, facebook, Twitter, Weibo, and forum discussion sites
  • Intent Mining - Mining user Intents or Intentions in Social Media Posts
    • Chapter 11 of my sentiment analysis book discusses this topic; also see this NAACL-HLT-2013 paper.
  • Generate a Book Automatically using Web Content: Synthesize knowledge on the Web (See our WWW-2003 paper).
    • Given a topic, discover sub-topics and related concepts, and synthesize information from multiple Web pages to present a coherent and fairly complete picture of the topic.
  • Web Data Extraction from Flat/Nested Data Records - Web page template detection and noise removal
    • Download MDR; Two new systems are DEPTA and NET.
  • Data Mining and Knowledge Discovery
    • Diagnostic Data Mining - Interestingness of Knowledge: Finding interesting and actionable knowledge.
      • We built a system in 2006, called Opportunity Map (also known as OM in Motorola). It was deployed in 2006 and is still in regular use in Motorola. Motorola has sold copies of the system to some Telco companies. There is also a patent, but unfortunately it is owned by Motorola, not us :-(.
    • Classification Based on Associations - Integrating Classification and Association Rule Mining
      • Download CBA - asociation rules based classification. You can also use the CBA implementation of Frans Coenen from Univesity of Livepool.
      • KDD-98 paper and others (KDD-99, PKDD-00, etc)
    • Association rule mining with multiple minimum supports
    • Interestingness: subjective interestingness, rule summary, rule query language, change mining

Recent Keynote and Invited Talks (not updated) -- (Some older talks)

1. Invited Talk. “Sentiment Analysis with Lifelong Learning.” ETS, December 7, 2015.
2. Invited Talk. “Sentiment Analysis with Lifelong Learning.” Brigham Young University, December. 3, 2015.
3. Keynote speech. “Sentiment Analysis, Lifelong Learning and Intelligent Personal Assistants.” The 2015 Conf. on Technologies and Applications of Artificial Intelligence (TAAI-2015). Taiwan, Nov. 20-22, 2015.
4. Invited talk. “Sentiment analysis and lifelong machine learning.” Frontiers in Computational Mathematics: AMS Central Fall Sectional Meeting, October 2-4, 2015.
5. Keynote speech. “The State of Sentiment.” Sentiment Analysis Symposium, New York City, July 15-16, 2015.
6. Invited tutorial. "Sentiment analysis: mining opinions, sentiments, and emotions." Sentiment Analysis Symposium, New York City, July 15-16, 2015.
7. Keynote speech. “Deception Detection via Pattern Mining of Web Usage Behavior” Workshop on Data mining for Big Data: Applications, Challenges & Perspectives, Morocco, 03/25/2015
8. Keynote speech. “Social Media Analysis via Continuous Learning.” Adobe Text Analytics Summit, Feb 26, 2015.
9. Some older talks

Teaching

• Spring 2019: CS 583 - Data Mining and Text Mining
• Fall 2018: CS 583 - Data Mining and Text Mining
• Spring 2018: CS 583 - Data Mining and Text Mining
• Fall 2017: CS 583 - Data Mining and Text Mining
• Spring 2017: CS 583 - Data Mining and Text Mining
• Fall 2016: CS 583 - Data Mining and Text Mining
• Fall 2015: CS 583 - Data Mining and Text Mining
• Spring 2015: CS 583 - Data Mining and Text Mining
• Past teaching

Professional Recognitions and Memberships

• ACM Fellow, Association for Computing Machinery (ACM).
• IEEE Fellow, Institute of Electrical and Electronics Engineers (IEEE).
• AAAI Fellow, Association for the Advancement of Artificial Intelligence (AAAI) (was Amercian Association of Artificial Intelligence).
• Member, ACM Special Interest Group on Knowledge Discovery and Data Mining (SIGKDD).

fake reviews: 2 and 3