def __init__(self, theNick, theRoom, theCommand, isPm=0):
Command.Command.__init__(self, theNick, theRoom, theCommand, isPm)
self.commands = {
"spam": {"method": BayesCommand.trainSpam, "acl": 1, "aclkey": Acl.TRAINSPAM},
"ham": {"method": BayesCommand.trainHam, "acl": 1, "aclkey": Acl.TRAINHAM},
"hammer": {"method": BayesCommand.addHammer, "acl": 1, "aclkey": Acl.HAMMER},
"spammer": {"method": BayesCommand.addSpammer, "acl": 1, "aclkey": Acl.SPAMMER},
"probability": {"method": BayesCommand.probability, "acl": 1, "aclkey": Acl.SPAMPROBABILITY},
}
self.spamFilter = SpamFilter.getInstance()
protocol = config.getString("login", "protocol")
#init the protocol we are going to use
Protocol.initProtocol( protocol )
protocol = Protocol.getProtocol()
#first get the cookie and id's
nick = config.getString( "login", "username" )
passwd = config.getString( "login", "password" )
events = Events.getEvents()
events.addListener( Display.getDisplay() )
events.addListener( SpamFilter.getInstance() )
#create the socket
sock = socket.socket( socket.AF_INET, socket.SOCK_STREAM )
sock.settimeout( 120.0 )
chatserver = config.getString( "login", "chatserver" )
chatport = config.getInt( "login", "chatport" )
#recreate the acl?
if sys.argv.count( "-newacl" ):
os.unlink( Acl.ACLFILE )
chatRoom = config.getString( "login", "chatroom" )
if len( sys.argv ) > 1:
# SPLITTING DATASET INTO TEST AND TRAINING SET -----------------------------------------
trainingSet = dataset[0:int(len(dataset) * TRAINING_SET_SIZE)]
testSet = dataset[int(len(dataset) * TRAINING_SET_SIZE):]
# LISTING TARGET VALUES AND ATTRIBUTES ----------------------------------------------
tv = ["ham", "spam"]
attributes = []
for i in range(0, len(trainingSet)):
attributes.extend([
x for x in dataset[i] if x != 'CLASS' if not x in tv
if not x in attributes
])
# TRAINING SPAM FILTER --------------------------------------------------------------
sf = SpamFilter.MultinomialSpamFilter(tv, attributes, trainingSet, 1)
sf.Learn()
# TESTING SPAM FILTER --------------------------------------------------------------
numTotalTries = len(testSet)
numTotalSuccesses = 0
falsePositives = 0 # when a ham is classified as spam
falseNegatives = 0 # when spam is classified as ham
realPositives = 0 # number of spam
realNegatives = 0 # number of ham
confusionMatrix = [[0, 0], [0, 0]] # rows are the real classes 0->HAM 1->SPAM
# columns are the stimated classes 0->HAM 1->SPAM
ajoin = [
"#mobsters", "#nova"
] #First channel in Array will be the active game channel, 2nd channel is the OPs channel
#Two bots are required. First bot will be the main game bot, while the second one is the missions bot
bots = [[rizon, Utility.Identity("mobsters", "mobster", None, ajoin)],
[rizon, Utility.Identity("DonVito", "dv", None, ajoin)]]
eng = Engine.Engine(event, bots, scheduler,
logger) #event, bots, sched, logger
eng.soft_start() # This is a mobster's mod
bots = eng.get_bots()
#(CORE) modules we are loading
scheduler.schedule_event(AuthSys.AuthSys(eng))
scheduler.schedule_event(KeepAlive.KeepAlive(eng))
event.add_mod(SpamFilter.SpamFilter(bots[0], eng)) #SpamFilter (mobsters only)
event.add_mod(Basic.Basic(eng))
## GAME WORK
#create the game instance
mobs = Mobster.Mobster(eng, bots[0], ajoin[0],
Security.Security(bots[0], ajoin))
msnclk = MissionClock.MissionClock(bots[1], mobs.host_channel)
regen = Regen.Regen(mobs, bots[0])
#Module Instances
event.add_mod(mobs) #Mobster Game (mobsters only)
event.add_mod(LegalAffairs.LegalAffairs(bots[0], mobs,
regen)) #Mobs (mobsters only)
event.add_mod(Missions.Missions(bots[1], mobs, msnclk)) #DonVito
# SPLITTING DATASET INTO TEST AND TRAINING SET -----------------------------------------
trainingSet = dataset[0: int(len(dataset)*TRAINING_SET_SIZE)]
testSet = dataset[int(len(dataset)*TRAINING_SET_SIZE):]
# LISTING TARGET VALUES AND ATTRIBUTES ----------------------------------------------
tv = ["ham","spam"]
attributes = []
for i in range(0,len(trainingSet)):
attributes.extend([x for x in dataset[i] if not x in tv if not x in attributes])
# TRAINING SPAM FILTER --------------------------------------------------------------
sf = SpamFilter.BinomialSpamFilter(tv,attributes,trainingSet)
sf.Learn()
# TESTING SPAM FILTER --------------------------------------------------------------
numTotalTries = len(testSet)
numTotalSuccesses = 0
falsePositives = 0 # when a ham is classified as spam
falseNegatives = 0 # when spam is classified as ham
realPositives = 0 # number of spam
realNegatives = 0 # number of ham