Skip to main content
When utilizing and applying various types of internal and external data,
the data must be prepared (preprocessed to make it usable).
Such preparations are said to account for approximately 80%
of the data utilization and application process, in terms of hours worked.
Hitachi’s Data Preparation Service analyzes data characteristics by using AI,
and supports the examination of data processing methods
as well as the verification of preprocessing logic in relation to such examinations.
The service reduces the burden of data preprocessing,
which is dependent on individual human skills, thereby enabling the customer
to rightfully focus on data analysis. It therefore helps customers utilize and apply their data.
Helping customers utilize their data
and supporting accelerated digital transformation
The service analyzes data specifications and data quality by using AI to support understanding of data such as item names, outliers, and relationships in the data. It reduces the burden of manual check work.
In addition to defined, standard processing methods (preprocessing logic), the service also streamlines the examination and verification of processing methods such as cleansing and integration by using functions that register and share expert knowledge and by using screens that do not need to be coded. The service reduces the burden of having to examine and verify data processing methods that are dependent on individual human skill and implemented skill.
The service enables customers to link verified data processing methods with ETL tools. The service promotes data utilization and application, by providing a seamless transition from data understanding to verification of data processing methods and actual operation.
Helping customers utilize and
apply their data through efficient data preprocessing
Data from stores and IoT data collected from factory equipment and other sources is abundant and diverse, meaning that data preprocessing for utilization and application often requires a large amount of time. This service detects unexpected data, and supports the examination of preprocessing logic, thereby reducing the burden of preprocessing and enabling the customer to focus rightfully on data analysis.
If the accumulated data (such as purchasing and usage results) to be utilized or applied includes outliers, the desired result might not be achieved. This service detects outliers and improves the quality of accumulated data before utilization and application. Consequently, in areas such as business expansion based on purchasing forecasts, and loss prevention based on detection of factory equipment breakdowns, the service leads to more precise utilization and application results.