Refactor how content is stored and validated

Right now, the CSV data is stored in a model 'Model'. Model has a list of SMBRows, which represent a CSV row. As the CSV file gets parsed, each row is stored as a string within SMBRow as 'String[] content'. This has made things very complicated since all data types (with different columns/types of columns) are all stored the same, as a list of strings. So if you want to reference a keyword row's match type, you would do something like getSMBRow().getContent()[3]. I propose that ads and keywords get stored in their own separate model, so we can reference them like getKeywordData().getMatchType(). This should make validating the data much easier.