Lingke: A Fine-grained Multi-turn Chatbot for Customer Service

Abstract

Traditional chatbots usually need a mass of human dialogue data, especially when using supervised machine learning method. Though they can easily deal with single-turn question answering, for multi-turn the performance is usually unsatisfactory. In this paper, we present Lingke, an information retrieval augmented chatbot which is able to answer questions based on given product introduction document and deal with multi-turn conversations. We will introduce a fine-grained pipeline processing to distill responses based on unstructured documents, and attentive sequential context-response matching for multi-turn conversations.

arXiv commentAccepted by COLING 2018 demonstration paper

References in corpus (4)

• Efficient Estimation of Word Representations in Vector Space
• Sequence to Sequence Learning with Neural Networks
• Improved Deep Learning Baselines for Ubuntu Corpus Dialogs
• Subword-augmented Embedding for Cloze Reading Comprehension

Cited by in corpus (6)

• DUMA: Reading Comprehension with Transposition Thinking
• A Neural Entity Coreference Resolution Review
• Topic-Aware Multi-turn Dialogue Modeling
• Towards End-to-End Open Conversational Machine Reading
• Open Vocabulary Learning for Neural Chinese Pinyin IME
• Effective Character-augmented Word Embedding for Machine Reading Comprehension