If you use Text Parsing or Text Miner nodes in SAS^® Text Miner, then the results might be incorrect. The problem occurs when a document contains characters that can be interpreted as HTML tags. The nodes might incorrectly remove text. As a result, some terms are missing from the analysis.

There is no general workaround to the problem.

There are no errors or warnings that indicate that text was removed incorrectly.

One specific example involves a less-than symbol (<) in a document. If a greater-than symbol (>) appears later in the document, then the nodes might incorrectly interpret the text between those two symbols as HTML. The text is stripped from the document.

To work around this specific example, remove less-than symbol / greater-than symbol pairs from your document. For example, use an editor to replace "<" with "lt", and to replace ">" with "gt".

Operating System and Release Information