The similarity index is the percentage of your text found, identically or very closely, in other sources. A high value is not automatically a problem: quotations and common expressions also count.
Focus on the highlighted blocks and the source list rather than the number alone. Two reports with the same percentage can differ greatly in severity.