NLPCC-ICCPOL 2016 - Call for Participation: Shared Tasks

The conference on Natural Language Processing and Chinese Computing (NLPCC) is the annual conference of CCF TCCI (Technical Committee of Chinese Information, China Computer Federation). The NLPCC conferences have been successfully held in Beijing (2012)，Chongqing (2013), Shenzhen (2014) and Nanchang (2015). This year’s NLPCC conference will be jointly held with The 24th International Conference on Computer Processing of Oriental Languages (ICCPOL) in Kunming on December 2 - 6, 2016.

NLPCC-ICCPOL 2016 will follow the NLPCC tradition of holding several shared tasks in natural language processing and Chinese computing. This year’s shared tasks focus on both classic problems and newly emerging problems, including Chinese word segmentation, question answering, word similarity measurement, stance detection, and sports news generation.

Participants from both academia and industry are welcomed. Each group can participate in one or multiple tasks and members in each group can attend the NLPCC-ICCPOL conference to present their techniques and results. A few selected groups will be invited to submit papers to the main conference and the accepted papers will appear in the conference proceedings published by Springer LNCS.

1. Overview of the Shared Tasks

There are five shared tasks in this year’s NLPCC conference and the details of each task can be found in the document of task guidelines. Here we give a very brief overview of each task.

◆ Chinese Word Segmentation for Weibo Text

This task aims to evaluate the techniques of Chinese word segmentation for Weibo texts. Since there exist different annotation criterions to segment Chinese words, these criterions should be considered to judge the performance of a word segmenter. Different with the traditional single evaluation measure for word segmentation, this task introduces a new criterion to evaluate the multi-granularity word segmentation.

◆ Open Domain Question Answering

This task aims to evaluate the open domain knowledge-based and document-based QA techniques in Chinese language.

◆ Chinese Word Similarity Measurement

This task provides a dataset of Chinese word similarity to evaluate and compare different semantic measures of lexical similarity, including 500 word pairs and their similarity scores.

◆ Stance Detection in Chinese Microblogs

This task is designed to evaluate stance detection techniques for Weibo texts, and stance detection aims to automatically determine whether the author of a Weibo text is in favor of the given target, against the given target, or neither. Noted here that the given target may not be present in the Weibo text. For example, for the given target of “全面放开二孩政策/Full liberalization of two-child policy”, in the Weibo text,“我们都是可以随便安排的机器么？/Are all of us easily arranged machines?”, the author’s stance is against the given target. This means that stance detection is different from traditional target (aspect)-dependent sentiment analysis.

◆ Sports News Generation from Live Webcast scripts

This task aims to evaluate document summarization techniques for producing Chinese sports news articles from live webcast scripts. The live webcast scripts are usually very long and the task is considered a special case of single document summarization.

2. How to Participate

Please fill out the registration form and send it to the coordinator Jiwei Tan(谭继伟) by email (NLPCCEval2016@163.com) before April 30, 2016.

If you have any question about the shared tasks, please do not hesitate to contact us by email.

3. Important Dates

◆ 2016/3/1：announcement of shared tasks and call for participation;

◆ 2016/3/31：release of detailed task guidelines;

◆ 2016/4/15：sample data release;

◆ 2016/4/30：registration deadline;

◆ 2016/6/30：test data release;

◆ 2016/7/5：systems’ results submission deadline;

◆ 2016/7/15：evaluation results release and call for system reports and conference papers;

◆ 2016/8/15：conference paper submission deadline;