The Impresso project develops application interfaces to facilate historical transmedia research through:
- the Impresso Web App, a user interface for content exploration and visualisation.
- the Impresso Datalab, a suite of tools for data exploration and analysis.
Specifically, the Impresso Datalab enables custom analyses of the Impresso corpus, and the semantic indexation of external document collections also with Impresso models. We offer access to the Impresso corpus, data and models via the Impresso Public API, a dedicated Python library, and HuggingFace. For more information, be sure to visit the Datalab website.
This repository contains notebooks that illustrate how to use the Impresso Public API (coming soon!) and Impresso Models, allowing you to search through Impresso data and use Impresso annotation models.
- Impresso Public API: The software component that provides third-party access to the Impresso backend.
- Impresso Python Library: The preferred method for users to interact with the Impresso Public API.
- Impresso Models: A collection of models trained to annotate the Impresso Corpus, made publicly available to facilitate the annotation of external documents, enabling comparison and analysis of semantic enrichments. Impresso models can be accessed through the Impresso Hugging Face organisation and via annotation services offered through the API.
Before getting started, check out how to create an account and obtain an API token on the Impresso Datalab website.
The notebooks in the starter folder will help you get started with the Impresso Public API and Python library:
The notebooks in the explore-vis folder help you build complementary views on your Impresso data:
The notebooks in the annotate folder demonstrate how to use Impresso models, either from the Hugging Face hub or through the Impresso API. These notebooks guide you in annotating your documents to produce annotations that are compatible with those in the Impresso corpus.
- Language Identification with impresso-pipelines Package
- OCR Quality Assessment with impresso-pipelines Package
- News Agencies Recognition and Linking with Impresso BERT models
Impresso - Media Monitoring of the Past is an interdisciplinary research project that aims to develop and consolidate tools for processing and exploring large collections of media archives across modalities, time, languages and national borders. The first project (2017-2021) was funded by the Swiss National Science Foundation under grant No. CRSII5_173719 and the second project (2023-2027) by the SNSF under grant No. CRSII5_213585 and the Luxembourg National Research Fund under grant No. 17498891.
Copyright (C) 2024 The Impresso team.
This program is provided as open source under the GNU Affero General Public License v3 or later.
