As the number of web pages on the internet grows, so do the conflicts between data gathered by search engines from different sources, primarily websites. Search engines require a conflict-resolving system that can quantify the truth of obtained data and assess the reliability of the information offered by sources. This paper presents a non-repetitive semantic approach for resolving data conflicts on search engines by including correctness elements into indexed data and estimating the trustwor-thiness of websites from which data are retrieved. Using the same dataset, our approach is validated in comparison with another approach based on several factors such as execution speed, information veracity efficiency and the error rate of sources’ trust. The approach shows promising results and outperforming the other approach. The execution time to process the dataset is about half an hour for our approach against four hours for the other one. The F1 measure of information truth calculations and the area under the ROC curve are around 88 and 94 percent, respectively. The source trustworthiness error is around 0.09 for our approach where is about 0.17 for the other one.