Real-world data in institutions is inconsistent and burdened by historical errors. We give experts tools that let them get an overview of it on their own — without programming skills and without waiting for the IT department.
What institutions come to us with
- Errors, empty titles or inconsistent records are searched for by hand, record by record.
- Every more complex query means a request to the system administrator.
- The original system contains the data, but search and statistics are slow and the interface does not support the necessary filter combinations or overviews.
- Adding the missing features directly to the original system would be technically risky or unnecessarily expensive.
How we work
- We create an analytical layer. We prepare a continuously updated copy of the required data, optimized for search, calculations and display. The original system remains the main place where data is created — we do not interfere with it.
- We set up rules and diagnostics. We define checks according to standards (MARC 21, MODS, NDK standards) and connect to authority and reference databases.
- We hand the interface over to the experts. Querying by logical conditions and text patterns, comparing sources, visualization and export of results.
Decision pattern
- What we automate: rule-based checks and comparisons with databases.
- What we prepare: overviews of problems and sets of records to resolve.
- What stays with the expert: decisions about the content of a record and its correction.
What you get
- An overview of your own data — fast analytical views, dashboards and statistics.
- Errors caught before they spread — checking whether records are ready for subsequent processes such as digitization.
- Less dependence on IT and fewer routine queries to administrators.
- Tools that are not tied to a single format — MARC 21 is key today, but libraries are gradually preparing for new data models.
Where it already works
Anakon allows librarians to work independently with metadata in MARC 21, MODS and the Kramerius 7 index, including diagnostics of typical digital library issues. The Kramerius statistics module collects and evaluates anonymized access data across installations — for reporting as well as for DNNT license statements. For digitization packages we check compliance with the standard using a validator, and for persistent identifiers we work with the CZIDLO tool (URN:NBN).
Transferable to: data quality control, data migrations and master data and code lists.
