Shuming Shi
Dr. Shuming Shi is a principal researcher of Tencent and a research director of Tencent AI Lab. His research interests include large language models, knowledge mining, natural language understanding, and natural language generation. He has published over 100 research papers in leading conferences and journals, such as ACL, EMNLP, AAAI, IJCAI, WWW, SIGIR, and TACL. He served as a co-chair of the EMNLP 2021 demonstration track, and served in the program committee of some conferences including ACL, EMNLP, WWW, AAAI, etc.
Demos & APIs & Datasets
- Effidit (Efficient and Intelligent Editing): A writing assistant with diversified and practical functions, including Text Completion, Error Correction, Text Polishing, K2S (Keyword-based sentence recommendation and generation), Cloud IME and etc. Specifically, text completion involves phrase completion, retrieval-based sentence completion and AI generation-based sentence continuation. Text polishing contains phrase-level polishing, sentence-level paraphrase and expansion.
[Online Demo | Details]
- TexSmart: A core NLP system that supports lexical, syntactic and semantic analysis of text in Chinese and English. In addition to the functions supported by a typical core NLP tool (such as word segmentation, POS tagging, and NER), TexSmart also provides some unique features: fine-grained NER, semantic expansion, and rich semantic represention of (certain types of) entities.
[Demo Link | Online HTTP API | Download offline toolkit]
- Tencent TranSmart: Translation-as-a-service via human-computer interaction, which aims to improve the translation efficiency of human translators with AI technologies.
[Demo Link | API (coming soon)]
- Large-Scale Chinese Word Embeddings: This corpus provides 200-dimension vector representations, a.k.a. embeddings, for over 8 million Chinese words and phrases, which are pre-trained on over 30 billion words.
Publications
Please refer to the Google Scholar page of Shuming Shi.
Links