claritystorm/cfpb-consumer-complaints

---
license: other
license_name: public-domain
task_categories:
- tabular-classification
- text-classification
- tabular-regression
tags:
- finance
- consumer-protection
- complaints
- nlp
- fintech
- united-states
pretty_name: CFPB Consumer Financial Complaints 2011-Present
size_categories:
- 10M<n<100M
---

# CFPB Consumer Financial Complaints 2011–Present

Every consumer financial complaint submitted to the Consumer Financial
Protection Bureau (CFPB) since 2011 — cleaned, structured, and ready
for analysis. 14M+ records covering credit reporting, debt collection,
mortgages, credit cards, banking, and more, with free-text consumer
narratives for 3.75M+ complaints.

**This repository contains a 1,000-row sample (Public Domain).**
Full dataset (CSV + Parquet) available at [claritystorm.com/datasets/cfpb-complaints](https://claritystorm.com/datasets/cfpb-complaints).

## Quick Start

```python
from datasets import load_dataset
ds = load_dataset("claritystorm/cfpb-consumer-complaints")

import pandas as pd
df = pd.read_csv("sample_1000.csv")
print(df["product_normalised"].value_counts())
print(df["submitted_via"].value_counts())
```

## Schema (selected fields)

| Field | Type | Description |
|-------|------|-------------|
| complaint_id | int | CFPB unique complaint identifier |
| date_received | string | Date CFPB received complaint (YYYY-MM-DD) |
| year | int | Year received (derived) |
| month | int | Month received (derived) |
| product | string | Financial product category (original) |
| product_normalised | string | Standardised product category |
| sub_product | string | Product sub-category |
| issue | string | Issue type |
| sub_issue | string | Issue sub-type |
| consumer_narrative | string | Consumer complaint text (free text, ~1.5M records) |
| has_narrative | int | 1 if narrative was provided, 0 otherwise |
| company | string | Company the complaint is about |
| state | string | US state (2-letter code) |
| zip_code | string | 5-digit ZIP code |
| submitted_via | string | Channel: Web / Phone / Referral / Postal mail / etc. |
| company_response | string | Company response to consumer |
| timely_response | int | 1 if company responded timely, 0 otherwise |
| consumer_disputed | int | 1 if consumer disputed company response |
| tags | string | Servicemember / Older American / Older American, Servicemember |

## Source

Consumer Financial Protection Bureau (CFPB) — Consumer Complaint Database.
CFPB data is a US federal government work in the **public domain** (17 U.S.C. 105).

Processed by [ClarityStorm Data](https://claritystorm.com).

---
许可证：其他
许可证名称：公有领域（Public Domain）
任务类别：
- 表格分类
- 文本分类
- 表格回归
标签：
- 金融
- 消费者权益保护
- 投诉
- 自然语言处理（Natural Language Processing，NLP）
- 金融科技（Fintech）
- 美国
展示名：CFPB消费者金融投诉数据集（2011年至今）
规模类别：样本量介于1000万至1亿条之间
---

# CFPB消费者金融投诉数据集（2011年至今）

本数据集收录2011年以来提交至美国消费者金融保护局（Consumer Financial Protection Bureau，CFPB）的全部消费者金融投诉，经清洗与结构化处理后可直接用于分析。数据集包含超1400万条记录，覆盖信用报告、债务催收、抵押贷款、信用卡、银行业务等多个金融领域，其中超375万条投诉附带消费者自由文本形式的投诉叙述内容。

**本仓库仅提供1000条数据的样本（公有领域授权）。完整数据集（CSV与Parquet格式）可于[claritystorm.com/datasets/cfpb-complaints](https://claritystorm.com/datasets/cfpb-complaints)获取。**

## 快速入门

python
from datasets import load_dataset
ds = load_dataset("claritystorm/cfpb-consumer-complaints")

import pandas as pd
df = pd.read_csv("sample_1000.csv")
print(df["product_normalised"].value_counts())
print(df["submitted_via"].value_counts())

## 字段架构（节选）

| 字段名 | 数据类型 | 字段说明 |
|-------|----------|----------|
| complaint_id | 整数 | CFPB专属投诉标识符 |
| date_received | 字符串 | CFPB接收投诉的日期（格式为YYYY-MM-DD） |
| year | 整数 | 投诉接收年份（派生字段） |
| month | 整数 | 投诉接收月份（派生字段） |
| product | 字符串 | 金融产品原始分类 |
| product_normalised | 字符串 | 金融产品标准化分类 |
| sub_product | 字符串 | 金融产品子类别 |
| issue | 字符串 | 投诉问题大类 |
| sub_issue | 字符串 | 投诉问题子类型 |
| consumer_narrative | 字符串 | 消费者自由文本投诉叙述（约150万条记录） |
| has_narrative | 整数 | 若消费者提供投诉叙述则为1，否则为0 |
| company | 字符串 | 被投诉的金融机构名称 |
| state | 字符串 | 投诉人所在美国州的2字母代码 |
| zip_code | 字符串 | 5位美国邮政编码 |
| submitted_via | 字符串 | 投诉提交渠道，包括网页、电话、转介、邮政邮件等 |
| company_response | 字符串 | 涉事机构对消费者的回复内容 |
| timely_response | 整数 | 若机构按时回复则为1，否则为0 |
| consumer_disputed | 整数 | 若消费者对机构回复提出异议则为1，否则为0 |
| tags | 字符串 | 标签类型，包括军人、老年消费者，或同时标注军人与老年消费者等 |

## 数据来源

数据来源为美国消费者金融保护局（CFPB）消费者投诉数据库。CFPB发布的数据属于美国联邦政府作品，处于**公有领域**（依据美国版权法17 U.S.C. 105条款）。本数据集由[ClarityStorm Data](https://claritystorm.com)整理加工。