Word Filterering

I need a word filtering system which creates smaller groups of keywords from a larger list based on filtering options. Example: You have a list of dogs say english setters dog irish setter dog irish setter dog food black lab dogs golden retrievers and so on So I could create groups for each breed and continue to filter down so first level group might be "irish setters" but within that group I could add another level for "food" which would contain "irish setter dog food". ## Deliverables 1. All phrases a. Dogs(dog) i. English setters dog(English setters) 1. English setter dog food(food)(chow) a. English setter dog wet food (wet) ii. Black lab dog(Black)(-spots) b. Cats(cat) i. Cat food(food) 1. Cat food with fish(fish) So this is an example of how the filters would work. The filter words are in bracket() and the an example of the phrase that could end up in the group is written before it. One thing is that it must go down levels within its own groups. So (food) is a filter only within its only group so cat food and English setter food are not combined even though they have the same filter. Also their can be more then one filter so the group that contains “english setter dog food?? would also contain “English setter chow?? as (food) and (chow) are both filters. Each group will contain hundreds of phrases especially at the top level so the (dog) group would have all types of dogs and other dog related phrases. However, their needs to be a way to only view the phrases that are left within the top level groups. So in this example if I was looking at the (dog) group then I would not see any (English setter) phrases as they have been moved into the group below it but I would see perhaps golden retriever dogs and be able to make a new group based on this. On top of the positive filters you must be able to put negative phrase on accounts. This is visible with the black lab dog. So the filter would be black but there is a negative filter of spots. So “black lab dogs?? would end up in this group but “dogs with black spots?? would not go in here and would remain up in the (dog) group. This process needs to be dynamic so if I add a new filter to a group then it should go back through the list of keywords above it and add the appropriate ones. Also if I add a negative filter it should move the phrases that are no longer suppose to be there out into the upper levels. So now that we have all these filters made up we should be able to dump more phrases in and they will go into the proper filter. So if I put in the phrase “English setter dog grooming?? it would be added to the 1.a.i group above along with the “English setters dog?? phrase.