The CKIP (Chinese Knowledge and Information Processing) group is a research team formed by the Institute of Information Science and the Institute of Linguistics of Academia Sinica in 1986. Its purpose is to establish a fundamental research environment for Chinese natural language processing. The preliminary goal of the project was to construct research infrastructures with reusable resources that could be shared by domestic and international research institutes. The accomplished resources include Chinese electronic dictionaries, Mandarin Chinese corpora, and processing technologies for Chinese texts. With these environments and technologies now well established, we are focusing on knowledge-based information processing. This area of research is motivated by the flood of information on the WWW for which effective and autonomous information processing tools are still lacking. To achieve high-level intelligent information processing, many of the most challenging research problems in the areas of knowledge acquisition, knowledge representation, and knowledge utilization are currently being addressed.
|Nov 2022 Our research paper — “HanTrans: An Empirical Study on Cross-Era Transferability of Chinese Pre-trained Language Model” has been accepted by “ROCLING 2022”.
|Jun 2022 Our research paper — “Converting the Sinica Treebank of Mandarin Chinese to Universal Dependencies” has been accepted by “LREC Workshop on LAW”.
|May 2022 Our research paper — “Conversational AI for Positive-sum Retailing under Falsehood Control” has been accepted by “ACL Workshop on NLP for Conversational AI”.
|Aug 2021 Our research paper — “H-FND: Hierarchical False-Negative Denoising for Distant Supervision Relation Extraction” has been accepted by “ACL findings”.
|Feb 2021 Our research paper — “Predict and Use Latent Patterns for Short-Text Conversation” has been accepted by “AAAI Workshop on DEEP-DIAL”.
Deep learning architectures such as deep neural networks, deep belief networks, recurrent neural networks and convolutional neural networks have been applied to fields including computer vision, speech recognition, natural language processing, audio recogn……
On knowledge representation area, we focus on the basic theory of knowledge ontology structure and the representation models for meticulous semantics. By analysis the nuance of synonyms, we found the representation method for meticulous semantics, and know……
We focus on concept-centric Chinese processing technology. The developed technology uses the statistics, language grammar, and common sense information obtained by automatic extraction as the basic knowledge to analyze the conceptual structure of the file ……
We research on how to automatically extract language knowledge and common sense. We expect that the language processing technology and the acquired knowledge can automatically analyze a large amount of text in the Internet and extract knowledge from it. Kn……