Curriculum Vitae

PUBLICATIONS:

2020:

17) Thanh Tran, Yifan Hu, Changwei Hu, Kevin Yen, Fei Tan, Kyumin Lee, and Se Rim Park, HABERTOR: An Efficient and Effective Deep Hatespeech Detector, EMNLP 2020 (full paper).

16) Thanh Tran, Di You, Kyumin Lee, Quaternion-based self-Attentive Long Short-term User Preference Encoding for Recommendation, CIKM 2020 research track (oral, 193/920 = 21%)

2019:

15) Yifan Hu, Changwei Hu, Thanh Tran, Tejaswi Kasturi, Elizabeth Joseph, Matt Gillingham, "What’s in a Name? – Gender Classification of Names with Character Based Machine Learning Models", Yahoo TechPulse (oral presentation, 49/685 = 7%).

14) Thanh Tran, Yifan Hu, Changwei Hu, Kevin Yen, Fei Tan, Se Rim Park, "Language Model for Hate-speech Classification in Yahoo News and Finance", Yahoo TechPulse (oral presentation, 49/685 = 7%).

13) Thanh Tran, Renee Sweeney, Kyumin Lee, "Adversarial Mahalanobis Distance-based Attentive Song Recommender for Automatic Playlist Continuation", SIGIR 2019 (acceptance rate 84/426 = 20%).

12) Thanh Tran, Xinyue Liu, Kyumin Lee, Xiangnan Kong, "Signed Distance-based Deep Memory Recommender", TheWebConf (or WWW) 2019 (acceptance rate 225/1247 = 18%).

2018:

11) Thanh Tran, Kyumin Lee, Yiming Liao, Dongwon Lee, "Regularizing matrix factorization with user and item embeddings for Recommendation", CIKM 2018 (acceptance rate 147/862 = 17%).

10) Thanh Tran, Madhavi Dontham, Jinwook Chung, Kyumin Lee, "Learning from Failure: How to Make a Failed Project Successful", under submission.

9) Thanh Tran, Madhavi R. Dontham, Kyumin Lee, "How to Succeed in Crowdfunding: a Long-Term Study in Kickstarter”, under submission.

2017:

8) Thanh Tran, Kyumin Lee, Nguyen Vo, Hongkyu Choi "Identifying On-time Reward Delivery Projects with Estimating Delivery Duration in a Crowdfunding Platform", ASONAM 2017 (acceptance rate 19%).

7) Thanh Tran, Kyumin Lee, "Characteristics of On-time and Late Reward Delivery Projects", ICWSM 2017.

6) Nguyen Vo, Kyumin Lee, and Thanh Tran, "MRAttractor: Detecting Communities from LargeScale Graphs", IEEE Big Data 2017 (acceptance rate 79/437 = 18%).

5) Yiming Liao, Thanh Tran, Dongwon Lee and Kyumin Lee, "Understanding Backing Patterns in Online Crowdfunding Communities", Websci 2017.

4) Nguyen Vo, Kyumin Lee, Cheng Cao, Thanh Tran and Hongkyu Choi "Revealing and Detecting Malicious Retweeter Groups", ASONAM 2017.

2016 and before:

3) Thanh Tran, Kyumin Lee, "Understanding Citizen Reactions and Ebola-Related Information Propagation on Social Media”, ASONAM, 2016.

2) Prudhvi Ratna Badri Satya, Kyumin Lee, Dongwon Lee, Thanh Tran and Jiasheng Zhang, "Uncovering Fake Likers in Online Social Networks", CIKM 2016.

1) Dac-Thanh TRAN, Duy-Hoa Ngo and Phan-Thuan DO, “An information content based partitioning method for the anatomical ontology matching task”, SoICT 2012.

RESEARCH EXPERIENCE:

Recommendation Systems (March 2017 – now):

• Exploit negative item embeddings to enrich recommendation results.

• Test our recommendation system on TasteProfile dataset (one-class setting), and MovieLens-10M, MovieLens-20M datasets (multiple-class setting), and show how our proposed model outperforms related existing methods.

Crowdfunding mining (March 2016 – March 2017):

• Mining 150,000 Kickstarter projects to predict their success rate and discover the temporal funding patterns of the successful projects.

• Understand distinguished investing patterns of crowdfunding investors and early detect investing patterns of investors.

• Understand the experience of successful creators in crowdfunding platforms by analyzing the project’s temporal backing pattern.

• Studying the delivery phase of the crowdfunding platforms. Then predicting which projects will deliver promised rewards on time, and suggest a more accurate reward delivery duration.

Large-Scale Distributed System : ZStore System - a distributed key-value storage system in ZingMe - the biggest social network in Vietnam (VNG, September 2011 – November 2011) :

• Researched cached algorithms and implemented ExpireLRU cache to ZStore.

• Researched hinted hand-off technique and integrated consistent hash to ensure fault-tolerance.

• Researched MapReduce technique to perform powerful queries.

• Researched NoSql storage systems : redis, cassandra, leveldb.

Social Computing : Friend Suggestion System - a system in which we suggested new friends for users (VNG, November 2011 - March 2012) :

• Researched and implemented topological features extracted from social graphs such as friendship graph, contact graph.

• Experimented Support Vector Machine as a learning method.

• Implemented model of MapReduce to calculate feature values faster.

INDUSTRY EXPERIENCE:

Wala, Jsc (April, 2012 - September 2014):

• Company Description: a startup company which lays research direction on online social network for mobile phone

• Position: R&D Engineer

• Project: Distributed Storage System, Profile Server, Feed Server, Channel Server, Redis Proxy

• Detail: Develop and maintain back-end system, establish services for social actions

Hanoi University of Science and Technology (June 2012 - August 2012):

• Position: Research Assistant

• Project: YAM++

• Details: Researched and implemented indexing and partitioning algorithms; design GUI for YAM++.

Hanoi University of Science and Technology (January 2012 - May 2012):

• Position: Research Assistant

• Project: An information content based partitioning method for the anatomical ontology matching task

• Details: Researched and proposed new approach to match anatomical ontologies

VNG Corporation (September, 2011 - March, 2012):

• Company Description: a leading corporation in popularizing social networks to Vietnamese.

• Position: Research Engineer

• Project:

- ZStore : a distributed key-value storage

Detail : Researched and improved key/value distributed storage systems.

- Friend Suggestion : A system which suggests a list of new friends for a certain user in ZingMe social network.

Detail: Researched and implemented Friend Suggestion System; improve performance of features querying by applying parallel processing and MapReduce technique.

- Feed Ranking : A system in which it ranks all feeds of an user to show his (her) feeds in descending order of important level.

Detail: implemented feed ranking system.

Luvina Jsc (September, 2010 - September, 2011):

• Company Description: a well-known Japan outsourcing company in VietNam

• Position: Software Engineer

• Project: C-Convert Project, Kyu-jin Project

• Detail: Convert C code from little endian to big endian in C-Convert Project; Coding and maintaining four sites. Two of them are on PC( http://job.nurse-senka.jp/, http://www.kaigojob.com/) and the remaining are on mobile (http://kj.m-sms.jp, http://mjob.nurse-senka.jp/).

Professional Services:

Talks:

• • Give a talk on MASR recommender at SIGIR 2019.

  • Present SDMR recommender system at 17th Annual Multidisciplinary Conference, Clark University, April 2019.

  • Present SDMR recommender system at WPI 50th anniversary of computer science department, 2019.

  • Give a talk on SDMR recommender at WWW 2019.

  • Give a talk on RME recommender system at CIKM 2018.

  • Give a talk on Ebola-related tweet mining on social media at ASONAM 2016.

Reviewer:

• • 2020: WWW, ICWSM.

  • 2019: JVLDB.

  • 2018: JVLDB.

External reviewer:

• • 2019: KDD, ICWSM, WWW.

  • 2018: KDD, WWW, AAAI.

  • 2017: KDD, WWW, SAC, ECIR, ECIR.

  • 2016: ECIR, SWDM, CIKM, ASONAM, TIST, ICWSM.

SOCIAL ACTIVITIES:

• • Volunteer at WWW 2019, SIGIR 2019.

  • WPI freshmen welcome, 2018.

  • ODSC East volunteer, 2018.

  • Trash collection on Logan canyon and Logan Wind Cave, 2016, 2017.

COMPUTER SKILLS:

• Programming: Python, Java, R.

• Deep learning framework: Pytorch.

• Platforms: Ubuntu.

• Databases: Relational Database: MySQL, SQLServer, NoSQL: Leveldb, Redis, Kyoto, Tokyo Cabinet, Memcached, CouchDB, BDB, Hadoop/Hbase, Cassandra

• Big data analytics frameworks: Hadoop, Spark.

HOBBIES:

Soccer, Music, Badminton, Reading.