ERIC Keyword Finder
Selecting appropriate keywords (or 'descriptors') is important for discoverability of research reports in the Camtree Digital Library. While you can select keywords manually, this tool can help you find relevant keywords from the ERIC Thesaurus based on the content of your report.
Instructions
- Upload your document (.pdf, .docx or .txt).
- Choose options: select 'semicolons/not capitalised' for Camtree reports. Double-pipes are useful for pasting lists into spreadsheets.
- Once the list has been generated, you can copy or download it, and edit it if required, adding domain-specific keywords. Remember that ERIC is US-based, so terminology and spelling reflects this.
The tool uses the Mistral Large LLM to suggest candidate keywords from the uploaded document, checks those suggestions against the ERIC Thesaurus to find exact or preferred descriptors, and uses sentence embeddings plus a similarity search to identify the nearest ERIC match when a suggested term is not in the thesaurus. Remember that LLM's can make mistakes! You should check that the generated abstract is correct for your purposes and edit it as necessary.