Issues with Explore tool results

We do our very best to support all types of URLs; both reference pages of patents and papers as well as the actual PDFs themselves. These URLs come in such a wide variety of formats and shapes, that occasionally - something can go wrong and the tool only picks up parts of the content. It can be the title, parts of the abstract, or even just some random text on the side of the page. It does not happen often, but when it does the contextual matching results are of course not going to be good. An indication of this issue could also be - if the text picked up is short - that the result list in the map is also short. If it contains 150 articles or less, this can indicate an issue with the input text.

Another issue could be the starting data set: Have you applied any filters to it? Remember that the tool only matches the input text to the active parts of your data set!

If you experience results that seem off, there’s a simple way to both validate whether this is the issue, and to get the results you were actually looking for.

• Double check that the data set you’re searching within does not have unintended filters.
• Rather than inputting the URL, copy the title of the paper/patent, and then the abstract of the document, to the Explore tool as free text.