From the World Wide Web to the Web of Data. The Web was originally designed as an information space, with the goal that it should be useful not only for human-human communication, but also that machines would be able to participate and help.
understood) by machines
This would be only possible if the Content of the information can be read and interpreted correctly. It means that it can be understood by machines. Only then machines can participate and help. Originally the way to deduce the meaning of the information was done by Natural Language Processing, which are utilized by Search Engines that try to extract the Meaning of the information.
Imagine what would be possible if the meaning of the information in the Web would not need to be extracted from the implicit information within Natural Language which is a error-prone process, but rather than before hand explicitly state what the information given really means.
The Semantic Web explicitly annotates the information Content of the Natural Language on the web with semantic meta-data. This semantic meta-data encodes the meaning of the Content which can be read and interpreted by machines correctly.
Thus, there are two ways, firstly via Natural Language Processing as it is done today and the other thing is besides encoding the information in Natural Language, give additional semantic annotation in an explicit way, so that the error prone process of extracting the Meaning by Natural Language Processing would be supported by already given explicit stated Semantic Annotation.