Artificial intelligence

Develop a spelling checker (i.e., best word predictor) using a 3-gram language model.

Each student needs to collect an Arabic corpus of 250k words at least. Tokenize the corpus into tokens/words, and then build a tri-gram language model for this corpus. The language model should contain: token, count, + the probability (or log) of the token, and should be saved in a CSV file. Develop an interface to allow the user to write text then click a spell button. If the user writes # in the text, the program will suggest the top five words (and their probability) as a replacement of the #, using the language model. The project should include. The source code, corpus, language model. The project should be JAVA.

please i need it as a code