Research project AG/GG/094 (Research action AG)
This project is a continuation of the Thesaurus project for the Belgostat Series of the National Bank of Belgium (Project AG/DD/36). Although this thesaurus significantly improves accessibility to the data sets not just for external users, but also to internal users, there is nevertheless still a gap in respect of accessibility to non-encoded and non-structured information, i.e. the metadata, articles, etc. Accordingly, the "text mining for the dissemination of statistical information via Belgostat on-line" project aims to provide users of the Bank’s website with fast and structured access to the methodological notes that underpin the compilation and production of the statistical data sets.
The introduction of "text mining", which will make searching for this type of information easier, is vitally important as regards offering users a comprehensive website complete with all necessary search engines.
On the basis of the inventory provided by the National Bank, the project will put forward a structure capable of simplifying the on-line dissemination of methodological notes via Belgostat (left, along with the data sets and tables, use of keywords, "text mining") as well as an adapted search and organisational method that takes account of the specific characteristics of the data.
The team of university specialists will start their activities after the Bank has drawn up an inventory of non-encoded information which users will be able to search for (list of methodological articles concerning the statistical series, metadata). Furthermore, this inventory will clarify the applicable design and maintenance rules.
The specialist team will produce a functional and technical proposal regarding the structure (arrangement, access, updating) of methodological notes, for use on the website of the bank and containing the following features:
- simple tracing of the non-encoded information
- links to the current thesaurus, making compatibility possible with statistical data sets.
- fine-tuning of use and control methods.
The Bank will incorporate this new application as part of its website, thereby supplementing the range of instruments being made available for information searches. The project will ensure simple access to methodological information concerning published data sets via the Statistical Website (Belgostat on-line).
