User documentation: Summarise documents#

Purpose#

This tool is used to automatically summarise documents using artificial intelligence. You upload a document and receive a structured summary according to your specifications. The system uses a large language model (LLM) to analyse the content and prepare it in various formats.

The basic principle is that the document is first converted into text form, then divided into processable sections, and finally summarised by the language model. For large documents, several partial summaries are created and synthesised into an overall result.

Range of functions#

The tool offers the following core functions:

  • Document processing: Extraction of text from PDF, DOCX and ODT files, including tables, headings and captions
  • Summary types: Six different output formats, from short versions to critical reflections
  • Multilingualism: Output of the summary in ten different languages
  • Individual instructions: Add your own specifications to customise the summary
  • Custom prompt: Complete control over the wording of the query to the language model

Operation#

Step-by-step instructions#

  1. Upload file: Select a file in PDF, DOCX or ODT format using the upload button
  2. Select summary type: Choose one of the following modes:

  • Short version: Compact overview with structural description

  • Detailed summary: Detailed analysis of all chapters and sections

  • Critical reflection: Evaluation highlighting positive and negative aspects

  • Thematic summary: Focus on core topics instead of chronological reproduction

  • Key points and references: Hierarchical presentation with complete bibliography

    • Own prompt only: Free formulation of the request
  1. Set language: Select the desired output language from the drop-down menu
  2. Optional - Additional instructions: Add specific requirements in the text field
  3. Start processing: The system analyses the document and generates the summary

Important notes#

  • The document must contain at least 200 characters of extractable text
  • Large documents are automatically divided into sections and processed step by step
  • Processing may take several minutes, depending on the size of the document
  • Always check the results critically, as AI language models are not error-free

Application example#

Initial situation#

You have a 30-page scientific paper in PDF format and would like to obtain a structured overview with a focus on the methodological aspects.

Procedure#

  1. Upload the PDF file
  2. Select ‘Detailed summary’
  3. Language: English
  4. Additional instruction: ‘Focus particularly on the methodological approaches and empirical results’
  5. Start processing

Result#

You will receive a structured summary in Markdown format, which first describes the overall context and then presents the content chapter by chapter, with particular emphasis on methodological aspects. The output is written in academic style and typically comprises 2-4 pages of text.

Recommendations for efficient use#

  • Precise instructions: Formulate additional instructions in a concrete and specific manner
  • Appropriate summary type: Select the mode according to your information needs
  • Document quality: Ensure that the document is well structured and readable
  • Iterative refinement: If the results are unsatisfactory, adjust the instructions and process again
  • Language consistency: For multilingual documents, it may be helpful to specify the document language in the instructions
  • Critical review: Compare summaries with the original document on a random basis

System limitations#

The tool CANNOT:

  • Interpret the content of images, graphics or diagrams (only captions are captured)
  • Process documents in formats other than PDF, DOCX and ODT
  • Guarantee the absolute accuracy and completeness of the summary
  • Present very long documents in a single summary without loss of information
  • Process handwritten or scanned documents without OCR pre-processing
  • Open password-protected or encrypted documents

Important limitations:

  • The summary is based exclusively on the extracted text
  • Very large documents may be shortened or simplified
  • The language model cannot draw on external sources for contextualisation
  • Complex formatting and layout structures may be lost

Summary#

The tool enables the efficient AI-supported summarisation of documents in various formats and languages. It is aimed at users who need a quick, structured overview of document content.

Your role is to provide suitable documents, select the appropriate summary type and critically review the results. The system supports you in processing information, but does not replace the need to engage with the content of the original document.

Backward Diagram Tool Flexible AI Mapping Forward