Task 2: Topic-Based Chinese Message Polarity Classification

Description :Topic-Based Chinese Message Polarity Classification: Given a message from Chinese Weibo Platform (Such as Sina, Tencent，NetEase etc. ) and a topic, classify whether the message is of positive, negative, or neutral sentiment towards the given topic. For messages conveying both a positive and negative sentiment towards the topic, whichever is the stronger sentiment should be chosen.

Each participant is required to submit two kinds of results based on: (1) restricted resource for fair comparison, e.g. same sentiment lexicon, corpus, etc. that will be announced together with the test data; and (2) unrestricted resource. We believe that a freely available, annotated corpus that can be used as a common testbed is needed in order to promote research that will lead to a better understanding of how sentiment is conveyed in tweets and texts.

Output format: We expect the following format for the output file:

runId <TAB> topicID <TAB> evalID<TAB> mesID <TAB>Polarity  

Where

runID is the ID of each participant run,   (runID is team name)

topicID is the ID of each topic,

evalID can be "0 for restricted test " or"1 for unrestricted test"，

mesID is message ID,

and Polarity can be predicted sentimentpolarity of topic ("1 for positive", "-1 for negative" or"0 for neutral").

For example, topic "iphone6"  ( TopicID 23),

<M15113801> 苹果公司已经发布了新产品iphone6 </M15113801>

<M15113803> iphone6运行速度快，还是不错的</M15113803>.

<M15113805> 但是，iphone6好像太薄了，容易折断，另外摄像头怎么是凸出的啊</M15113805>

Output file:

NA <TAB> 23 <TAB>1 <TAB> M15113801 <TAB> 0

NA <TAB> 23 <TAB>1 <TAB> M15113803 <TAB> 1       

NA <TAB> 23 <TAB>1 <TAB> M15113805 <TAB> -1

Training Data

Test input data for Task 2 is now available here. (Fixed some data bugs including ambiguous annotations, missed topicID etc.)

Test data

Test data for Task 2 is now available here.

Registration

Please send the signed ‍‍license form‍‍ to Liao Xiangwen (liaoxw@fzu.edu.cn)

Co-organizers

Liao Xiangwen，Fuzhou University

Xu Ruifeng, Harbin Institute of Technology

Li Binyang, University of International Relation

Xu Liheng, Institute of Automatic, CAS

Committee

Xu Hongbo, Institute of Computing Technology, CAS

Wang Suge, Shanxi University

Huang Minlie, Tshinghua University

Liu Kang, Institute of Automatic, CAS

Wan Xiaojun, Peking University

Wang Houfeng, Peking University

Lin Hongfei, Dalian University of Technology

Liu Yiqun, Tshinghua University

Zhang Qi, Fudan University

Xu Weiran, Beijing University of Post Technology

Yao Tianfang, Shanghai Jiaotong University

Importance Dates

Registration for Bakeoffs open: 2015-02-01
Training data released: 2015-03-10
Registration for Bakeoffs close: 2015-04-01
Dry run (format validation): 2015-05-01
Test data released: 2015-05-11 (18:00 Beijing Time)
Test result submission deadline: 2015-05-13 (18:00 Beijing Time)
Test result evaluation released: 2015-05-15
Technical report submission deadline: 2015-06-01
Technical report reviews returned: 2015-06-11
Camera-ready submission deadline: 2015-06-21
Workshop date: 2015-07-30/31