Corpus - how to build the body
"Corpus" is a Latin word which means "body". If it sounds like "corpse" (dead body), well, that is because it literally got its meaning from the concept of dead body and later on it branched out into many fields, including linguistics, music, literature, religion, and so on. What they all have in common is the meaning: "corpus" means "body".
Our evidence service needs a body to perform its magic. As we saw in the previous section, the TF-IDF factor cannot be calculated if we don't have body made out of hundreds of articles. So lets begin by creating a corpus.
Lets go back to the selected news item from the previous section.
Lets say we want to see if our application can find and organize enough evidence related to that news title and provide us some insight. One way to build the corpus around this article is to find articles related to important keywords in the original news item. Checking the contents of that news, we can find the following names and keywords...