48. Amazon FinSpace-CSDN博客

本文链接：https://blog.csdn.net/meiyubaihe/article/details/121389742

Overview

Amazon FinSpace is a fully managed data management and analytics service that makes it easy to store, catalog, and prepare financial industry data at scale, reducing the time it takes for financial services industry (FSI) customers to find and access all types of financial data for analysis from months to minutes.
With FinSpace, you can collect data in a secure data management application and catalog it by relevant business concepts such as asset class, risk classification, or geographic region, which makes it easy to discover and share across your organization.
FinSpace supports your organization’s compliance requirements by enforcing data access controls and keeping audit logs.
Financial services customers store petabytes of data which is collected from both internal and external data sources.

Import data easily — The SDKs allows you to load data files into FinSpace in bulk, daily, or ad-hoc fashion. Connect your daily historical data feeds from stock exchanges and data providers into FinSpace.
Store and catalog data with business terms — Create a business data catalog with your business taxonomy to organize data so that your business users can easily discover it. Organize data by asset classes, regions, data types, or industry.
Track versions of data — Create bi-temporal views that let you analyze data the way it looked at a particular date and time. Reproduce historical financial models for audit and compliance purposes.
Prepare and analyze data at scale — Use FinSpace notebook with integrated managed Spark clusters to run analysis on petabytes of data. Scale compute with spark clusters on an as-needed basis
Financial time series analysis — Run financial time series analysis on high density market data using integrated time series library with over 100 embedded functions including statistical and technical indicators such as Bollinger Bands.

Launch FinSpace from your AWS console, and configure how data will be organized in the catalog for easy searching.
Add data that will be needed for analytics.
Organize and describe the data so that it can be searched from the catalog.
Prepare data by creating historical or current data views partitioned to optimize performance.
Analyze data using integrated Jupyter notebooks and managed Spark clusters for data processing at scale.

You can ingest data into FinSpace from your enterprise data lake or on-premises data stores.
The data is ingested into Datasets.
FinSpace supports ingestion of structured data such as CSV, parquet, XML, and JSON or any unstructured data files
Dataset is a logical container of semantically identical data and schema.
Data is ingested into a dataset as a changeset, and every time a new set of data is added to a dataset, a changeset is created.

Datasets can be described, organized, and made browsable and searchable in FinSpace.
The organizational concepts provided in Finspace are designed to provide centralized governance and control.
Categories
- Categories allow for cataloging of datasets by commonly used business terms.
- Categories are hierarchical in nature, allowing for each node of the hierarchy to have a name and a description.
- The order of the nodes within a level are defined when you define categories.
Controlled Vocabularies
- Controlled Vocabularies are enumeration lists of attributes to describe datasets.
- A controlled vocabulary is used to ensure that standardized terms are used to describe a dataset
Attribute Sets
- Attribute Sets are lists of attributes that can be applied to describe datasets.
- Attributes are metadata fields used to capture additional business context for each dataset
- You can then browse and search attributes to find a dataset based on the values assigned to the attributes.