Paragraph search

The semantic search of the INTERGATOR Patent Search allows the entry of individual terms, but also of paragraphs or entire documents. Especially the exemplary search based on sentence-by-sentence text sections is a decisive advantage of the artificial intelligence-based patent search. In contrast to the use of individual search terms, which are sometimes imprecise or ambiguous, the context of a complete sentence is more likely to provide a context of content and thus a more precise paraphrase for the search. So instead of starting a search with individual terms, one can copy a complete paragraph into the search box and the AI will evaluate from the total set of words contained and their relationship to each other and thus more relevant search hits. 

This alternative search approach is made possible by the artificial intelligence behind it, which no longer searches only for the explicit occurrence of the search term within the documents, but understands each data record in its thematic entirety. Which terms occur in the document? How often does the term occur? What are core terms in the document and how does the document relate to similar content and the entire data set? This evaluation makes each document unique and locates it thematically. In this way, each dataset is given a digital fingerprint that makes it unique. The automatic categorization and the extension of the content search to a context of meaning also makes it possible to find similar hits that are thematically similar but use completely different terms. Especially in patent literature searches, the inclusion of equivalent paraphrases is more challenging. However, since INTERGATOR Patent Search does not focus exclusively on the mere occurrence of hits, similar but relevant search hits are also found.