QUASI

QUASI数据集由南加州大学构建，包含4,699个问题及其对应的24,006个答案句子，旨在解决多答案场景下的答案整合问题。数据集通过从Quora等来源收集问题，并由人工从网络中检索相关答案句子，再通过众包工人进行筛选和分组，确保每个问题下的答案覆盖不同方面。QUASI数据集特别适用于智能助手和搜索引擎等应用，帮助系统提供既全面又简洁的用户响应。

The QUASI dataset, developed by the University of Southern California, contains 4,699 questions and their corresponding 24,006 answer sentences, and is designed to tackle the problem of answer consolidation in multi-answer scenarios. The dataset collects questions from sources such as Quora, has relevant answer sentences retrieved from the web manually, and then employs crowdworkers to screen and group these answers, ensuring that the answers under each question cover diverse aspects. The QUASI dataset is particularly suitable for applications like intelligent assistants and search engines, helping these systems provide comprehensive yet concise user responses.