Finance Commons and the Bad Data Toolbox
Ready to use document AI toolbox, optimized for bad data
- OCronos: OCR correction decoding model used to correct OCR errors.
- Segmentext: A text segmentation encoding model used to improve text structure.
- Bibteker: Structured Literature Information Extraction Encoding Model.
- Pleias Editor: Integrated process to make bad text suitable for advanced retrieval applications.
- Reverse Zotero: A tool that automatically converts unstructured books into BiBTex data.
- Support the generation of synthetic data that is close to actual production use, in order to develop more robust LLM and embedding models.
Product Details
Finance Commons and Bad Data Toolbox are a series of models and tools for document AI research and application. They focus on handling bad data, including OCR errors, disorganized text, etc., to improve the robustness of AI in document processing. These tools and models help automate processes, reduce the workload of preparing content for businesses, and support the development of the next generation of multimodal document models.
