Abstract:This paper describes our participation in STC-2 Chinese subtask of NTCIR-13. All runs are submitted for both two tasks, namely generation-based task and retrieval-based task. Various methods based on Seq2Seq framework were used to generate the responses. Interaction-focused method based on deep learning models is used to deal with relevance between queries and comments. As for generation-based task, we propose to add a constraint term to the original objective function of Seq2Seq, which further incorporates the coherent between post and reply into consideration. Specifically, three different approaches have been adopted to measure the similarity between post and reply, i.e. unlearned constraint such as cosine function and pretrained matching function such as bilinear function and MatchPyramid. As for information retrieval task, we apply a method to obtain relevance that is crucial to retrieval. As called interactionfocused method, we apply method based on MatchPyramid, which first builds interactions between a query and a comment and then uses a deep model to obtain the representation for the interactions and the relevance score.

Overview of the NTCIR-10 INTENT-2 Task.

Overview of the NTCIR-9 INTENT Task.

Summary of the NTCIR-10 INTENT-2 task

Summary Of The Ntcir-10 Intent-2 Task: Subtopic Mining And Search Result Diversification

Overview of the NTCIR-11 IMine Task.

Overview of the NTCIR-12 IMine-2 Task

THUIR at NTCIR-10 INTENT-2 Task

THUIR at NTCIR-9 INTENT Task.

THUSAM at NTCIR-11 IMine Task

Overview of the ntcir-14 we want web task

Overview of the NTCIR-15 We Want Web with CENTRE (WWW-3) Task

Overview of the ntcir-13 we want web task

THUIR at NTCIR-12 IMine Task.

Information Access Technologies

Overview of NTCIR-13

THUIR at the NTCIR-14 WWW-2 Task.

Mining Search Subtopics from Query Logs.

Gbot at the NTCIR-13 STC-2 Task Hainan

Microsoft Research Asia at NTCIR-12 STC Task.

Report on NTCIR-13

ISCAS at NTCIR-3: Monolingual, Bilingual and MultiLingual IR Tasks