Write a method loadDictionary which reads all the words in the dictionary file into a suitable data structure. Since each word is on a separate line, with no leading or trailing whitespace, you could use the readLine method from the BufferedReader class for reading words into the dictionary. You can find a simple example of input and output here. You will need to decide for yourself which data-structure to use. There are a number of possibilities, e.g. a simple array, a binary search tree, a hashtable etc. The important consideration is to that whatever data-structure you choose provides an efficient lookup mechanism. Remember that the two dictionary files are in alphabetical order so care would be needed if, for example, you used a binary search tree. The simplest approach is probably to use an array for storage, with binary search for lookup, remembering that the words in the dictionary files are in alphabetical order.
Now open the text file and read each word in turn. This needs some care. You will have to decide how a word should be defined for present purposes. Remember also that you are required to provide the line number in the file for any word that is misspelled. This means keeping track of what line you are on. (My own solution reads the text file line by line, making it easy to get the line number; I would then ideally use a StringTokenizer to obtain the words, except that its way of setting delimiters is inconvenient, so I wrote my own simple StringParser. But there are several ways of doing this.)