客服会话总结摘要数据

对合法采集的企业服务会话数据进行分析、处理，帮助企业快速对会话内容进行概括总结，提升客服服务效率。客服服务用户过程中，因技能组差异，可能需要转交客服，被转交客服为了了解用户的问题，需要理解以前客服和用户的交互信息，通过总结的摘要和用户核心诉求等关键信息，有助于客服快速理解以往的服务记录，提升服务效率。客服会话总结摘要数据的算法规则包括：
1、数据采集和处理：企业客户自身积累的客服和用户之间的语聊会话数据，对用户的id等敏感信息做匿名化处理。
2、算法加工：客服和用户间的语聊会话存在大量数据，抽取部分样本并人工总结摘要及抽取核心诉求，以此为样本，微调训练 NLP大模型。针对每日新增的客服和用户的语聊数据，基于前述的场景数据微调的NLP大模型，生成会话摘要，及用户的核心诉求等字段关键信息。

This dataset analyzes and processes legally collected enterprise service session data to help enterprises quickly summarize session content and improve customer service efficiency.
During customer service interactions with users, transfers between representatives may be required due to differences in skill sets. When a user is transferred to a new customer service agent, the agent needs to understand the previous interactions between the original customer service and the user. Key information such as the summarized session abstract and the user's core demands can help the new agent quickly grasp the past service records, thereby improving service efficiency.
The algorithmic specifications for generating customer service session summary data are as follows:
1. Data Collection and Processing: Anonymize sensitive information such as user IDs from the voice chat session data between customer service staff and users accumulated by enterprise clients themselves.
2. Algorithmic Processing: There is a large volume of voice chat session data between customer service and users. We extract a subset of samples, manually summarize their session abstracts and extract core user demands, and use these samples to fine-tune an NLP-based large language model. For the daily newly added voice chat session data between customer service and users, the fine-tuned NLP large language model (based on the aforementioned scenario-specific data) is used to generate key information fields including session summaries and user core demands.

搜集汇总

数据集介绍

特点

该数据集包含客服与用户的会话数据及其总结摘要，每日更新，规模达999999条。主要用于提升客服效率，通过NLP大模型生成会话摘要和核心诉求，帮助企业快速理解用户需求。