Wenbo Wang @ Kno.e.sis Center

Research Experience

• Understanding and Modeling Emotions with Tweets (2011.11-present)

  There is nothing more exciting than embracing the era of of big data. We aim to study people's emotions at the level of millions of data entries. We plan to collect, analyze, model emotions in social media and eventually predict people's emotions by the texts people write. Here are some preliminary but interesting discoveries on Twitter users from Estern Standard Timezone (US & Canada) between Nov. 10th and Nov. 28th:
  • The most significant (more than 65%) emotion on Nov. 24th is thankfulness. (of course^_^) diagram link
  • The most significant emotion after people get up is thankfulness (see the peak between 7am and 9am); The most significant emotion at night is love(see the love/affection peak around 10pm). diagram link

• Unsupervised Sentiment Clue Extraction in Social Media(2011.6-present)

  Computers are good at numbers, while humans are good at words. When it comes to sentiment analysis, besides presenting sentiment polarities (numbers between -1(negative) and 1(positive)), we extract meaningful sentiment clues (both words and phrases). We think it is more interesting and convincing to tell a user that a movie is "must see" and "rate 5 starts" than to say the overall polarity is 0.85. By applying an optimization model to minimize the inconsistency relations among sentiment clues, we extract target-dependent word and phrase clues as well as corresponding polarities from tweets.

• Discover Fine-grained Emotions in Suicide Note(2011.6-2011.9)

  Suicide takes away over 98 lives each day in US [stat]. If we can identify and keep track of people's emotions, we will be able to monitor people's mental status and prevent people from committing suicide. We constructed a hybrid classifier that is able to discover the sentence-level emotions from 16 categories, e.g., love, pride, abuse, anger, happiness, guilt, etc. We explored a variety of lexical, syntactic and knowledge-based features. we also proposed an algorithm to automatically extract effective syntactic and lexical patterns from training examples.

• Twitris++: 360° Social Media Analysis(2011.6-present)

  Twitris is a Semantic Web application that facilitates understanding of social perceptions by Semantics-based processing of massive amounts of event-centric data. My contributions include:
  • Designed and implemented a multi-process program to collect tweets for multiple events, collect user profiles and decode user locations to latitudes and longitudes.
  • Improved website response time by optimizing database schema, MySQL database and SQL queries

• Internship in WOO (A Web Of Objects) Team @ Yahoo! (2010.6-2011.4)

  The goal of WOO project is to integrate existing high quality knowledge bases, such as DBpedia, Yahoo! Movies, etc. and come with an integrated WOO knowledge base. My job is to disambiguate and integrate movie, actor and director objects in DBpedia and Yahoo! Movies. My contributions include:
  • Designed and implemented object disambiguation algorithm in Hadoop environment
  • Generic features as well as domain/problem specific features
  • Automatically collected training data for supervised machine learning

• Human Performance Cognition Ontology(HPCO) (2008.9-2010.6)

  The ultimate goal of HPCO is to better perform knowledge discovery by semantic search and browsing. Focused domain hierarchy is semi-automatically constructed from Wikipedia and triples are extracted from scientific literatures (PubMed). My job is to align predicates in extracted triples with existing predicates in domain ontology and the idea is to automatically discover verb relationships like, synonym/antonym, from seed verb synonyms/antonyms. My contributions include:
  • Utilized 100 CPU cores to process large biomedical paper abstracts over Hadoop DFS
  • Extracted a restricted set of seed synonym/antonym verbs from Wordnet
  • Constructed probability enabled synonym/antonym patterns from POS tagged training corpus
  • Applied learned patterns to obtain more synonym/antonym verbs

• Research of Services intelligence in Convergence Network Environment @ BUPT(2005-2008)

  • Extended OWL-S profile to describe telecom service more precisely
  • Implemented a Semantic Parlay X Gateway

• RoboCup Competition(CSU_Yunlu Team) @ CSU(2004-2005)

  • Proposed and implemented an integrated proactive information update model
  • Proposed and implemented opponent strategy modeling

Education

• 2008-present PhD student in Computer Science Department at Wright State University (GPA: 4.0) Advisor: Professor Amit Sheth


• 2005.9-2008.4 Master student in School of Computer Science and Technology at Beijing University of Posts and Telecommunications (BUPT), China. Advisor: Professor Xiaofeng Li


• 2001.9-2005.6 Bachelor student in School of Information Science and Engineering at Central South University (CSU), China. Advisor: Professor Jun Peng

Publications

• Wenbo Wang, Lu Chen, Ming Tan, Shaojun Wang, Amit P. Sheth. Discovering Fine-grained Sentiment in Suicide Notes. Biomedical Informatics Insights, 2012 (to appear)
• Lu Chen, Wenbo Wang, Meenakshi Nagarajan, Shaojun Wang and Amit P. Sheth. Beyond Positive/Negative Classification: Automatic Extraction of Sentiment Clues from Microblogs. Kno.e.sis Center Technical Report 2011
• Wenbo Wang, Christopher Thomas, Amit Sheth, Victor Chan. Pattern-Based Synonym and Antonym Extraction. 48th ACM Southeast Conference, ACMSE2010, Oxford Mississippi, April 15-17, 2010
• Christopher J. Thomas, Wenbo Wang, Pankaj Mehra, Delroy Cameron, Pablo N. Mendes, and Amit P. Sheth.. What Goes Around Comes Around – Improving Linked Opend Data through On-Demand Model Creation. In: Proceedings of the WebSci10: Extending the Frontiers of Society On-Line, April 26-27th, 2010, Raleigh, NC: US.
• Ashutosh Jadhav, Wenbo Wang, Raghava Mutharaju, Pramod Anantharam, Vinh Nyugen, Amit P. Sheth, Karthik Gomadam, Meenakshi Nagarajan, and Ajith Ranabahu, Twitris: Socially Influenced Browsing, Semantic Web Challenge 2009, demo at 8th International Semantic Web Conference, Oct. 25-29 2009, Washington, DC, USA

Awards

• 2006 Second place in BEA IMS Application Innovation and Development Competition
• 2005 Eleventh place in Osaka Qualification for International RoboCup 2005 2D Simulation
• 2004 Second prize of Hunan Province in China Mathematics Contest in Modeling

Courses

• IHE601 Engineering Academic Integrity

• MTH607 Optimization Techniques

• CS712 Advanced Topics in AI
• CS875 Semantic Web

• CS714 Machine Learning I
• CS790 Knowledge Rep for Semantic Web

• CS675 Web Information Systems(Audit)
• CS784 Programming Languages
• CS699 Cloud Computing

• CEG730 Distributed Computing Principles

• CEG633 Operating Systems
• CS701 Database Systems and Design
• CS790 Web 3.0 Social Services and Sensor Semantic Web

• CS707 Information Retrieval
• CS740 Computational Complexity and Algorithm Analysis

• CS610 Theoretical Foundations of Computing
• CS875 Semantic Web

Computer Skills

• Programming Languages: Java, C, Scheme, PHP, XML, HTML, SQL
• Web Technologies: RDF, OWL, SPARQL, XML, HTML
• Operating Systems: Windows, Linux/Unix

Resume

Contact Information

wenbo AT knoesis DOT org

Wenbo Wang

Joshi 380(Semantic Web Lab)

Kno.e.sis Center

Wright State University

3640 Colonel Glenn Hwy.

Dayton, Ohio 45435-0001