UNLV-ShaneD
commented
12 years ago Possible solution: Grouping of transactions into items can be done similarly to how Gmail filters messages. We can have an ordered list of filters that will automatically sort transactions into their respective items. Additionally, being able to preview and edit what items freshly scraped transactions will be assigned to is an elegant way to approach this.
Note that we have the following information available to us for each transaction: -Amount -Description -Category (May not be provided by the online statement being scraped)
Filtering can be accomplished simply by using regular expressions on the description. If the category field is available, filtering may also be done using that. Obviously, for user-friendliness, we may want to have a simple sub-string search option as another type of filter.
The user interface for grouping transactions into items can be particularly challenging due to the large amount of data that needs to be processed. Having the user go through each and every transaction individually is clearly not an option.
We need a smart interface for grouping transactions into items.
For clarity, the current glossary entry for "item" is attached: item A series of financial transactions and their associated data (amounts, descriptions) - all of which are based on a single recurring or non-recurring expense. Example: Automotive fuel is a (recurring and seasonal) item. Not an example: Credit card payments may seem like an item, but they actually mix and hide different recurring items below (gas, food, etc.) - they are not an actual expense. Credit card interest, however, may be an item.