Deleting documents from a Lucene index is easy when you have a primary key field in your document (like in traditional SQL databases).
However, sometimes deleting a number of documents based on multiple fields in the document is what you need. The Lucene API allows you to achieve this by specifying a query to use for deletion.
To do this, pick the right Analyzer, construct the query, pass the query to the indexWriter to delete the documents.
It's not immediately obvious, but the analyzer that you are using makes a huge difference to the way your query is run. This is because the StandardAnalyzer filters out common English words like "the" and "a". You might want to pick a different analyzer (like KeywordAnalyzer) so that it matches exactly. This obviously depends on you application of Lucene of course.