def Item_based():
user_id=session['user_id'][0]
productData = functions.flipPersonToPlaces(reviewdata1.reviews)
print "Finding similar Places " # for single place
similar_item=functions.mostSimilar(productData,"4iTRjN_uAdAb7_YZDVHJdg")
print dict(similar_item)
b_data=[]
for bid in dict(similar_item).keys():
sql = ("SELECT B_NAME, PHOTO_URL,RATING from BUSINESS_CA where B_ID = '%s'" % bid)
cursor.execute(sql)
data = cursor.fetchone()
if data != None:
b_data.append([data[0],data[1],data[2]])
print b_data
print "******************"
print "Computing Item Similarity" # for entire product data
itemSimilarity = functions.computeItemSimilarities(productData)
print itemSimilarity
print "******************"
print "Item Based Filtering for Recommendations"
recommendedplc_ib=functions.itemBasedFiltering(reviewdata1.reviews,str(user_id),itemSimilarity)
print recommendedplc_ib.keys() #send to shivani
locations =[]
for bid in recommendedplc_ib.keys():
sql = ("SELECT B_NAME, LATITUDE, LONGITUDE, ADDRESS, RATING from BUSINESS_CA where B_ID = '%s'" % bid)
cursor.execute(sql)
data_map = cursor.fetchone()
if data_map!=None:
value=[str(data_map[0]),float(data_map[1]),float(data_map[2]),str(data_map[3]),data_map[4]]
locations.append(value)
print "locations is"
print locations
return render_template('ItemRecommend.html',location=locations,similarplaces=b_data)
def user_based():
user_id=session['user_id'][0]
print "Most similar reviewers/Users "
similar_user=functions.mostSimilar(reviewdata1.reviews,str(user_id))
#similar_user=functions.mostSimilar(reviewdata.reviews,"T9hGHsbJW9Hw1cJAlIAWmw")
u_data = []
print dict(similar_user).keys()
print "###########"
for uid in dict(similar_user).keys():
sql = ("SELECT USER_NAME from USER where USER_ID = '%s'" % uid)
cursor.execute(sql)
data = cursor.fetchone()
if data != None:
rating= random.randint(3,5)
u_data.append([data[0],rating])
print "U_DATA"
print u_data
print "###########"
print "Place Recommendations for a user"
recommendplc_ub=functions.getRecommendations(reviewdata1.reviews,str(user_id)) #how much will one user like a particular place
print "************"
locations =[]
for bid in recommendplc_ub.keys():
sql = ("SELECT B_NAME, LATITUDE, LONGITUDE, ADDRESS, RATING from BUSINESS_CA where B_ID = '%s'" % bid)
cursor.execute(sql)
data_map = cursor.fetchone()
if data_map!=None:
value=[str(data_map[0]),float(data_map[1]),float(data_map[2]),str(data_map[3]),data_map[4]]
locations.append(value)
return render_template('UserRecommend.html', location=locations,similarusers=u_data)
