SwissText is an annual conference established in 2016 by the Zurich University of Applied Sciences (ZHAW) that brings together text analytics experts from industry and academia. KONVENS has been held regularly since 1992 and deals with all areas of natural language processing. This year and for the first time, the two conferences will be held jointly and due to the COVID-19 situation online.
On June 25, we will present a HIBU-based search solution by DSwiss (Zurich): high-security digital safes which allow users to store, exchange, but also search any type of documents and other security-relevant data.
The focus will be on the text analytics aspects of the solution developed with HIBU. Since the uploaded data can contain any sort of content, the solution supports users to organize their data in two ways: by a hierarchical folder structure and by means of facets (search filters). Some of the default facets are derived from structured metadata as file format or date, while others are populated dynamically by semantic taggers and classifiers as e.g. semantic document type, persons, locations mentioned in the document. Especially these filters have proven very useful to support document and data retrieval.
We touch on the challenges of analyzing and indexing documents in a highly secure, multiple-encrypted environment and will then discuss joint ongoing work to support the individual needs of users even better:
- use state-of-the-art neural network architectures to classify and extract more types of information from documents to provide a broader range of filters;
- personalize the trained models that create the search filters; and
- support smart actions with text-based triggers (e.g. proposing a specific folder when uploading a document).
For a sneak peak of this talk, please read our customer success story “Fully encrypted intelligent search for digital safes“.