北京航空航天大学大数据高精尖中心自然语言处理研究团队开展了智能问答的研究与应用总结。包括基于知识图谱的问答（KBQA），基于文本的问答系统（TextQA），基于表格的问答系统（TableQA）、基于视觉的问答系统（VisualQA）和机器阅读理解（MRC）等，每类任务分别对学术界和工业界进行了相关总结。

本SDK为实现QQ机器人平台的大一统，支持多种机器人平台：酷Q、MyPCQQ、QQLight等

Vega-Lite Chart Dataset and NL Generation Framework using LLMs

Official PyTorch implementation for ״ lassification-Regression for Chart Comprehension״

NS-CQA: the model of the JWS paper 'Less is More: Data-Efficient Complex Question Answering over Knowledge Bases.' This work has been accepted by JWS 2020.

本SDK为实现QQ机器人平台的大一统，支持多种机器人平台：酷Q、MyPCQQ、QQLight等

MARL: the model of the IJCAI 2020 paper 'Retrieve, Program, Repeat: Complex Knowledge Base Question Answering via Alternate Meta-learning.' This work has been published by IJCAI 2020.

Implementation of "Distilling Knowledge for Fast Retrieval-based Chat-bots" (SIGIR 2020) using deep matching transformer networks and knowledge distillation for response retrieval in information-seeking conversational systems.

QDEE: Question Difficulty and Expertise Estimation in Community Question Answering Sites

SoftSVM implementation

A code quality assurance tool for Unity 3D.

Predicting Future Collaborations in Community-based Question Answering (CQA) Networks

Sample code to extract different types of users in CQA in particular context

Code Samples Validation (v2 Dashboard) - automated testing with centralized reporting based on GitHub Actions, Playwright and Docusaurus