Please go to http://openresearch.baidu.com/topic/71.jspx for detailed description and registration.

 Notice: Submission time is from [2014-07-15 00:00:00] to [2014-10-01 00:00:00]. Time is based on PDT(UTC-7)。

Welcome to the 2014 CIKM Competition!

The 2014 CIKM Competition welcomes you to compete on the real datasets that has never been released before the public. Fresh from CIKM competition committee, the task is about to classify the unlabeled queries into certain catagories by using the given labeled training dataset . We hope that this task will challenge you intellectually and push the limits of data-mining capabilities.

From Baidu, we encourage you use these data fully for your algorithmic innovations and computational discoveries, and make breakthrough discoveries in science and industrial problems

We hope you enjoy the competition, and wish you great success!

Background

Contemporary users are no longer satisfied with the search results that only literally match their queries. They expect the search engine to cognitively understand the implied meaning of their queries and return the information accordingly. Therefore, query detection (QD) technology has emerged as an integral part of all major commercial search engines. QD technology enables the search engines to understand the most probable query intent based on their query logs. For instance, take Baidu, the world's largest Chinese search engine, as an example. When a user searches using the query “Beijing Shanghai”, the search engine will not simply return the text snippets containing the two cities names. Instead, based on its query log, the search engine understands that the most probable query intent is to find information on “travelling from Beijing to Shanghai”, and generates the search results accordingly.