bla1221212/Bitext-retail-banking-llm-chatbot-training-dataset

--- license: cdla-sharing-1.0 task_categories: - question-answering - table-question-answering language: - en tags: - question-answering - llm - chatbot - banking - conversational-ai - generative-ai - natural-language-understanding - fine-tuning - retail-banking pretty_name: >- Bitext - Retail Banking Tagged Training Dataset for LLM-based Virtual Assistants size_categories: - 10K<n<100K --- # Bitext - Retail Banking Tagged Training Dataset for LLM-based Virtual Assistants ## Overview This hybrid synthetic dataset is designed to be used to fine-tune Large Language Models such as GPT, Mistral and OpenELM, and has been generated using our NLP/NLG technology and our automated Data Labeling (DAL) tools. The goal is to demonstrate how Verticalization/Domain Adaptation for the [Retail Banking] sector can be easily achieved using our two-step approach to LLM Fine-Tuning. For example, if you are [ACME Bank], you can create your own customized LLM by first training a fine-tuned model using this dataset, and then further fine-tuning it with a small amount of your own data. An overview of this approach can be found at: [From General-Purpose LLMs to Verticalized Enterprise Models](https://www.bitext.com/blog/general-purpose-models-verticalized-enterprise-genai/) The dataset has the following specifications: - Use Case: Intent Detection - Vertical: Retail Banking - 26 intents assigned to 9 categories - 25545 question/answer pairs, with approximately 1000 per intent - 1224 entity/slot types - 12 different types of language generation tags The categories and intents are derived from Bitext's extensive experience across various industry-specific datasets, ensuring the relevance and applicability across diverse banking contexts. ## Dataset Token Count The dataset contains a total of 4.98 million tokens across 'instruction' and 'response' columns. This extensive corpus is crucial for training sophisticated LLMs that can perform a variety of functions including conversational AI, question answering, and virtual assistant tasks in the banking domain. ## Fields of the Dataset Each entry in the dataset comprises the following fields: - flags: tags - instruction: a user request from the Retail Banking domain - category: the high-level semantic category for the intent - intent: the specific intent corresponding to the user instruction - response: an example of an expected response from the virtual assistant ## Categories and Intents The dataset covers a wide range of banking-related categories and intents, which are: - **ACCOUNT**: check_recent_transactions, close_account, create_account - **ATM**: dispute_ATM_withdrawal, recover_swallowed_card - **CARD**: activate_card, activate_card_international_usage, block_card, cancel_card, check_card_annual_fee, check_current_balance_on_card - **CONTACT**: customer_service, human_agent - **FEES**: check_fees - **FIND**: find_ATM, find_branch - **LOAN**: apply_for_loan, apply_for_mortgage, cancel_loan, cancel_mortgage, check_loan_payments, check_mortgage_payments - **PASSWORD**: get_password, set_up_password - **TRANSFER**: cancel_transfer, make_transfer ## Entities The entities covered by the dataset include: - **{{Full Name}}**, typically present in intents such as apply_for_loan, apply_for_mortgage. - **{{Banking App}}**, featured in intents like activate_card, check_loan_payments. - **{{Account Number}}**, relevant to intents such as activate_card_international_usage, block_card. - **{{Customer Support Working Hours}}**, associated with intents like customer_service, human_agent. - **{{Customer Support Team}}**, important for intents including cancel_card, make_transfer. - **{{Company Website URL}}**, typically present in intents such as activate_card, apply_for_loan. - **{{Customer Support}}**, featured in intents like activate_card, block_card. - **{{Customer Support Email}}**, relevant to intents such as activate_card_international_usage, apply_for_loan. - **{{Mortgage Account Number}}**, associated with intents like cancel_mortgage, check_mortgage_payments. - **{{Mortgage Account}}**, important for intents including check_loan_payments, check_mortgage_payments. - **{{Billing}}**, typically present in intents such as check_fees, check_mortgage_payments. - **{{Username}}**, featured in intents like activate_card, block_card. - **{{Customer Support Phone Number}}**, relevant to intents such as activate_card, apply_for_loan. - **{{Live Chat}}**, associated with intents like activate_card_international_usage, apply_for_mortgage. - **{{Company Website}}**, important for intents including activate_card, apply_for_loan. - **{{Mortgage Department}}**, typically present in intents such as apply_for_mortgage, cancel_mortgage. - **{{Account}}**, featured in intents like activate_card, block_card. - **{{Name}}**, relevant to intents such as activate_card, apply_for_loan. - **{{Bank Name}}**, associated with intents like activate_card, apply_for_loan. - **{{Password}}**, important for intents including activate_card, block_card. - **{{Customer Support Email Address}}**, typically present in intents such as activate_card, apply_for_loan. - **{{Customer Service Email Address}}**, featured in intents like activate_card_international_usage, cancel_card. - **{{Email Address}}**, relevant to intents such as activate_card, apply_for_loan. - **{{Profile}}**, associated with intents like cancel_card, check_fees. - **{{Customer Service Working Hours}}**, important for intents including activate_card, apply_for_loan. - **{{Credit Card}}**, typically present in intents such as activate_card, block_card. - **{{Bank App}}**, featured in intents like activate_card, block_card. - **{{Loan Account Number}}**, relevant to intents such as cancel_loan, check_loan_payments. - **{{Account Settings}}**, associated with intents like activate_card, block_card. This comprehensive list of entities ensures that the dataset is well-equipped to train models that are highly adept at understanding and processing a wide range of banking-related queries and tasks. ## Language Generation Tags The dataset includes tags indicative of various language variations and styles adapted for Retail Banking, enhancing the robustness and versatility of models trained on this data. These tags categorize the utterances into different registers such as colloquial, formal, or containing specific banking jargon, ensuring that the trained models can understand and generate a range of conversational styles appropriate for different customer interactions in the retail banking sector. ## Language Generation Tags The dataset includes tags that reflect various language variations and styles, crucial for creating adaptable and responsive conversational AI models within the banking sector. These tags help in understanding and generating appropriate responses based on the linguistic context and user interaction style. ### Tags for Lexical variation - **M - Morphological variation**: Adjusts for inflectional and derivational forms in banking terminology. - Example: "is my account active", "is my account activated" - **L - Semantic variations**: Handles synonyms, use of hyphens, compounding common in banking communications. - Example: “what's my balance date", “what's my billing date” ### Tags for Syntactic structure variation - **B - Basic syntactic structure**: Simple, direct commands or statements. - Example: "activate my card", "I need to check my balance" - **I - Interrogative structure**: Structuring sentences in the form of questions. - Example: “can you show my balance?”, “how do I transfer money?” - **C - Coordinated syntactic structure**: Complex sentences coordinating multiple ideas or tasks. - Example: “I want to transfer money and check my balance, what should I do?” - **N - Negation**: Expressing denial or contradiction. - Example: "I do not wish to proceed with this transaction, how can I stop it?" ### Tags for language register variations - **P - Politeness variation**: Polite forms often used in customer service. - Example: “could you please help me check my account balance?” - **Q - Colloquial variation**: Informal language that might be used in casual customer interactions. - Example: "can u tell me my balance?" - **W - Offensive language**: Handling potentially offensive language which might occasionally appear in frustrated customer interactions. - Example: “I’m upset with these charges, this is ridiculous!” ### Tags for stylistic variations - **K - Keyword mode**: Responses focused on keywords relevant to banking tasks. - Example: "balance check", "account status" - **E - Use of abbreviations**: Common abbreviations in the context of banking. - Example: “acct for account”, “trans for transaction” - **Z - Errors and Typos**: Includes common misspellings or typographical errors found in customer inputs. - Example: “how can I chek my balance” ### Other tags not in use in this Dataset - **D - Indirect speech**: Expressing commands or requests indirectly. - Example: “I was wondering if you could show me my last transaction.” - **G - Regional variations**: Adjustments for regional language differences. - Example: American vs British English: "checking account" vs "current account" - **R - Respect structures - Language-dependent variations**: Formality levels appropriate in different languages. - Example: Using “vous” in French for formal addressing instead of “tu.” - **Y - Code switching**: Switching between languages or dialects within the same conversation. - Example: “Can you help me with my cuenta, please?” These tags not only aid in training models for a wide range of customer interactions but also ensure that the models are culturally and linguistically sensitive, enhancing the customer experience in retail banking environments. ## License The `Bitext-retail-banking-llm-chatbot-training-dataset` is released under the **Community Data License Agreement (CDLA) Sharing 1.0**. This license facilitates broad sharing and collaboration while ensuring that the freedom to use, share, modify, and utilize the data remains intact for all users. ### Key Aspects of CDLA-Sharing 1.0 - **Attribution and ShareAlike**: Users must attribute the dataset and continue to share derivatives under the same license. - **Non-Exclusivity**: The license is non-exclusive, allowing multiple users to utilize the data simultaneously. - **Irrevocability**: Except in cases of material non-compliance, rights under this license are irrevocable. - **No Warranty**: The dataset is provided without warranties regarding its accuracy, completeness, or fitness for a particular purpose. - **Limitation of Liability**: Both users and data providers limit their liability for damages arising from the use of the dataset. ### Usage Under CDLA-Sharing 1.0 By using the `Bitext-retail-banking-llm-chatbot-training-dataset`, you agree to adhere to the terms set forth in the CDLA-Sharing 1.0. It is essential to ensure that any publications or distributions of the data, or derivatives thereof, maintain attribution to the original data providers and are distributed under the same or compatible terms of this agreement. For a detailed understanding of the license, refer to the [official CDLA-Sharing 1.0 documentation](https://cdla.dev/sharing-1-0/). This license supports the open sharing and collaborative improvement of datasets within the AI and data science community, making it particularly suited for projects aimed at developing and enhancing AI technologies in the retail banking sector. --- (c) Bitext Innovations, 2024license: CDLA共享1.0（CDLA-Sharing 1.0） task_categories: - 问答 - 表格问答 language: - en tags: - 问答 - 大语言模型（LLM） - 聊天机器人 - 银行业 - 对话式AI - 生成式AI - 自然语言理解 - 微调 - 零售银行业 pretty_name: Bitext——面向基于大语言模型的虚拟助手的零售银行标注训练数据集 size_categories: - 10K<n<100K # Bitext——面向基于大语言模型的虚拟助手的零售银行标注训练数据集 ## 概述 本混合合成数据集专为微调大语言模型（Large Language Model）设计，适配模型包括GPT、Mistral及OpenELM，由我方自然语言处理（Natural Language Processing, NLP）/自然语言生成（Natural Language Generation, NLG）技术与自动化数据标注（Data Labeling, DAL）工具生成。其核心目标是展示，借助我方提出的大语言模型微调两步法，可轻松实现零售银行业领域的垂直化适配/领域自适应。举例而言，若您是ACME银行，可先通过本数据集训练微调模型，再使用少量自有数据对其进行二次微调，从而打造专属定制化大语言模型。该方法的详细概述可参阅：《从通用大语言模型到垂直化企业模型》（链接：https://www.bitext.com/blog/general-purpose-models-verticalized-enterprise-genai/） 本数据集规格如下： - 用例：意图检测 - 垂直领域：零售银行业 - 9大类别下涵盖26项意图 - 共25545条问答对，单意图平均约1000条 - 1224种实体/槽位类型 - 12种语言生成标签类型 本数据集的类别与意图均基于Bitext在多行业专属数据集领域的丰富经验构建，确保其在各类银行业务场景中具备相关性与适用性。 ## 数据集Token统计 本数据集的"instruction"与"response"两列总计包含498万个Token。该大规模语料库对于训练具备复杂功能的大语言模型至关重要，可使其在银行业领域胜任对话式AI、问答及虚拟助手等各类任务。 ## 数据集字段 本数据集每条数据包含以下字段： - flags：标签 - instruction：零售银行业领域的用户请求 - category：意图对应的高级语义类别 - intent：与用户指令对应的具体意图 - response：虚拟助手的预期回复示例 ## 类别与意图 本数据集涵盖多类银行业相关类别与意图，具体如下： - **ACCOUNT（账户）**：check_recent_transactions（查询近期交易）、close_account（注销账户）、create_account（开立账户） - **ATM（自动取款机）**：dispute_ATM_withdrawal（ATM取款纠纷）、recover_swallowed_card（取回被吞卡片） - **CARD（卡片）**：activate_card（激活卡片）、activate_card_international_usage（激活卡片国际使用功能）、block_card（冻结卡片）、cancel_card（取消卡片）、check_card_annual_fee（查询卡片年费）、check_current_balance_on_card（查询卡片当前余额） - **CONTACT（联系客服）**：customer_service（客户服务）、human_agent（人工客服） - **FEES（费用查询）**：check_fees（查询费用） - **FIND（查找网点/设备）**：find_ATM（查找ATM）、find_branch（查找营业网点） - **LOAN（贷款）**：apply_for_loan（申请贷款）、apply_for_mortgage（申请抵押贷款）、cancel_loan（取消贷款）、cancel_mortgage（取消抵押贷款）、check_loan_payments（查询贷款还款情况）、check_mortgage_payments（查询抵押贷款还款情况） - **PASSWORD（密码）**：get_password（获取密码）、set_up_password（设置密码） - **TRANSFER（转账）**：cancel_transfer（取消转账）、make_transfer（发起转账） ## 实体 本数据集涵盖的实体如下： - **{{Full Name}}（全名）**：通常出现在申请贷款、申请抵押贷款等意图中 - **{{Banking App}}（银行应用）**：常见于激活卡片、查询贷款还款情况等意图 - **{{Account Number}}（账户号码）**：适用于激活卡片国际使用功能、冻结卡片等意图 - **{{Customer Support Working Hours}}（客服工作时间）**：关联客户服务、人工客服等意图 - **{{Customer Support Team}}（客服团队）**：对取消卡片、发起转账等意图至关重要 - **{{Company Website URL}}（公司网站URL）**：通常出现在激活卡片、申请贷款等意图中 - **{{Customer Support}}（客服支持）**：常见于激活卡片、冻结卡片等意图 - **{{Customer Support Email}}（客服邮箱）**：适用于激活卡片国际使用功能、申请贷款等意图 - **{{Mortgage Account Number}}（抵押贷款账户号码）**：关联取消抵押贷款、查询抵押贷款还款情况等意图 - **{{Mortgage Account}}（抵押贷款账户）**：对查询贷款还款情况、查询抵押贷款还款情况等意图至关重要 - **{{Billing}}（账单）**：通常出现在查询费用、查询抵押贷款还款情况等意图中 - **{{Username}}（用户名）**：常见于激活卡片、冻结卡片等意图 - **{{Customer Support Phone Number}}（客服电话号码）**：适用于激活卡片、申请贷款等意图 - **{{Live Chat}}（在线聊天）**：关联激活卡片国际使用功能、申请抵押贷款等意图 - **{{Company Website}}（公司网站）**：对激活卡片、申请贷款等意图至关重要 - **{{Mortgage Department}}（抵押贷款部门）**：通常出现在申请抵押贷款、取消抵押贷款等意图中 - **{{Account}}（账户）**：常见于激活卡片、冻结卡片等意图 - **{{Name}}（姓名）**：适用于激活卡片、申请贷款等意图 - **{{Bank Name}}（银行名称）**：关联激活卡片、申请贷款等意图 - **{{Password}}（密码）**：对激活卡片、冻结卡片等意图至关重要 - **{{Customer Support Email Address}}（客服电子邮箱）**：通常出现在激活卡片、申请贷款等意图中 - **{{Customer Service Email Address}}（客服服务电子邮箱）**：常见于激活卡片国际使用功能、取消卡片等意图 - **{{Email Address}}（电子邮箱）**：适用于激活卡片、申请贷款等意图 - **{{Profile}}（个人档案）**：关联取消卡片、查询费用等意图 - **{{Customer Service Working Hours}}（客服工作时间）**：对激活卡片、申请贷款等意图至关重要 - **{{Credit Card}}（信用卡）**：通常出现在激活卡片、冻结卡片等意图中 - **{{Bank App}}（银行应用）**：常见于激活卡片、冻结卡片等意图 - **{{Loan Account Number}}（贷款账户号码）**：适用于取消贷款、查询贷款还款情况等意图 - **{{Account Settings}}（账户设置）**：关联激活卡片、冻结卡片等意图 本实体列表覆盖全面，可确保数据集能够充分训练模型，使其熟练理解并处理各类银行业相关查询与任务。 ## 语言生成标签 本数据集包含各类语言变体与风格标签，适配零售银行业场景，可提升基于本数据集训练的模型的鲁棒性与通用性。这些标签将用户话语划分为不同语域，如口语化、正式或包含特定银行行话，确保训练后的模型能够理解并生成适配零售银行业不同客户交互场景的多样化对话风格。 ### 词汇变体标签 - **M - 词形变化（Morphological variation）**：适配银行术语的屈折与派生形式 示例："is my account active"、"is my account activated" - **L - 语义变体（Semantic variations）**：处理同义词、连字符使用及银行沟通中常见的复合词 示例："what's my balance date"、"what's my billing date" ### 句法结构变体标签 - **B - 基础句法结构（Basic syntactic structure）**：简单直接的命令或陈述 示例："activate my card"、"I need to check my balance" - **I - 疑问句式结构（Interrogative structure）**：以疑问句形式组织语句 示例："can you show my balance?"、"how do I transfer money?" - **C - 并列句法结构（Coordinated syntactic structure）**：包含多个观点或任务的复合句 示例："I want to transfer money and check my balance, what should I do?" - **N - 否定表达（Negation）**：表达否认或矛盾 示例："I do not wish to proceed with this transaction, how can I stop it?" ### 语域变体标签 - **P - 礼貌程度变体（Politeness variation）**：客户服务中常用的礼貌表达 示例："could you please help me check my account balance?" - **Q - 口语化变体（Colloquial variation）**：休闲客户交互中可能使用的非正式语言 示例："can u tell me my balance?" - **W - 冒犯性语言（Offensive language）**：处理客户沮丧时可能出现的冒犯性语言 示例："I’m upset with these charges, this is ridiculous!" ### 风格变体标签 - **K - 关键词模式（Keyword mode）**：聚焦于银行业务相关关键词的回复 示例："balance check"、"account status" - **E - 缩写使用（Use of abbreviations）**：银行业场景中常见的缩写 示例："acct for account"、"trans for transaction" - **Z - 错误与拼写错误（Errors and Typos）**：包含客户输入中常见的拼写错误或笔误 示例："how can I chek my balance" ### 本数据集未使用的其他标签 - **D - 间接引语（Indirect speech）**：间接表达命令或请求 示例："I was wondering if you could show me my last transaction." - **G - 区域变体（Regional variations）**：适配区域语言差异 示例：美式英语与英式英语差异："checking account" vs "current account" - **R - 敬语结构（Respect structures）**：依赖于语言的形式程度变化 示例：法语中使用"vous"进行正式称呼而非"tu" - **Y - 代码切换（Code switching）**：同一场对话中切换语言或方言 示例："Can you help me with my cuenta, please?" 这些标签不仅有助于训练模型适配各类客户交互场景，还可确保模型具备文化与语言敏感性，从而提升零售银行业环境中的客户体验。 ## 许可证 本数据集`Bitext-retail-banking-llm-chatbot-training-dataset`采用**社区数据许可协议（Community Data License Agreement, CDLA）共享1.0**版本发布。该许可协议旨在促进广泛的共享与协作，同时确保所有用户使用、共享、修改及利用该数据集的自由不受侵犯。 ### CDLA-Sharing 1.0的核心条款 - **署名与相同方式共享**：用户必须对本数据集进行署名，并将衍生作品以相同许可协议进行共享 - **非排他性**：本许可协议为非排他性，允许多个用户同时使用本数据集 - **不可撤销性**：除存在实质性违约情况外，本许可协议下的权利不可撤销 - **无担保**：本数据集按"现状"提供，不提供任何关于其准确性、完整性或特定用途适用性的担保 - **责任限制**：用户与数据提供方均不对因使用本数据集产生的损害承担责任 ### CDLA-Sharing 1.0协议下的使用 使用`Bitext-retail-banking-llm-chatbot-training-dataset`数据集即表示您同意遵守CDLA-Sharing 1.0协议的条款。任何对本数据集或其衍生作品的发布或分发，都必须保留对原始数据提供方的署名，并采用本协议或兼容的许可条款进行分发。 如需详细了解本许可协议，请参阅[CDLA-Sharing 1.0官方文档](https://cdla.dev/sharing-1-0/)。 本许可协议支持AI与数据科学社区内数据集的开放共享与协作改进，尤其适用于旨在开发和提升零售银行业领域AI技术的项目。 --- (c) Bitext Innovations, 2024