Hi Erwin,Marcus,
I need a framework where i can override the topic selection for responses based on psychodynamics theories . My plan is to apply defense mechanisms (explained in psychodynamics) to understand the state of mind of the chatter & choose a topic accordingly. For ex, sublimation is a defense mechanism wherein instinctual impulses are converted to socialize activities like sports, art, competition . When this state of mind is detected, a response is selected which are not closely related to or reminds of our instinctual thoughts like food, sleep, sex etc. Likewise, if the user input talks about buying a new car or dress, this can be matched to “undoing” defense mechanism.
Basically, i want to parse the input text to match it to a defense mechanism & then find a response matching the defense mechanism.
Does the above sound sensible?
vasanth