API tools faq
paste
ivanakarpuzova

userItemSist

ivanakarpuzova
Sep 10th, 2019
78
0
Never
Add comment
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
text 10.04 KB | None | 0 0
raw download clone embed print report
  1. from math import sqrt
  2.  
  3. oceniPoKorisnici = {
  4. 'Lisa Rose': {'Catch Me If You Can': 3.0, 'Snakes on a Plane': 3.5, 'Superman Returns': 3.5,
  5. 'You, Me and Dupree': 2.5, 'The Night Listener': 3.0, 'Snitch': 3.0},
  6. 'Gene Seymour': {'Lady in the Water': 3.0, 'Snakes on a Plane': 3.5, 'Just My Luck': 1.5, 'The Night Listener': 3.0,
  7. 'You, Me and Dupree': 3.5},
  8. 'Michael Phillips': {'Catch Me If You Can': 2.5, 'Lady in the Water': 2.5, 'Superman Returns': 3.5,
  9. 'The Night Listener': 4.0, 'Snitch': 2.0},
  10. 'Claudia Puig': {'Snakes on a Plane': 3.5, 'Just My Luck': 3.0, 'The Night Listener': 4.5, 'Superman Returns': 4.0,
  11. 'You, Me and Dupree': 2.5},
  12. 'Mick LaSalle': {'Lady in the Water': 3.0, 'Snakes on a Plane': 4.0, 'Just My Luck': 2.0, 'Superman Returns': 3.0,
  13. 'You, Me and Dupree': 2.0},
  14. 'Jack Matthews': {'Catch Me If You Can': 4.5, 'Lady in the Water': 3.0, 'Snakes on a Plane': 4.0,
  15. 'The Night Listener': 3.0, 'Superman Returns': 5.0, 'You, Me and Dupree': 3.5, 'Snitch': 4.5},
  16. 'Toby': {'Snakes on a Plane': 4.5, 'Snitch': 5.0},
  17. 'Michelle Nichols': {'Just My Luck': 1.0, 'The Night Listener': 4.5, 'You, Me and Dupree': 3.5,
  18. 'Catch Me If You Can': 2.5, 'Snakes on a Plane': 3.0},
  19. 'Gary Coleman': {'Lady in the Water': 1.0, 'Catch Me If You Can': 1.5, 'Superman Returns': 1.5,
  20. 'You, Me and Dupree': 2.0},
  21. 'Larry': {'Lady in the Water': 3.0, 'Just My Luck': 3.5, 'Snitch': 1.5, 'The Night Listener': 3.5}
  22. }
  23.  
  24.  
  25.  
  26.  
  27.  
  28. movies = {
  29. 'Lady in the Water': {'Gary Coleman': 1.0, 'Larry': 3.0, 'Jack Matthews': 3.0, 'Michael Phillips': 2.5,
  30. 'Gene Seymour': 3.0, 'Mick LaSalle': 3.0, 'Michelle Nichols': 3.0},
  31. 'Snakes on a Plane': {'Jack Matthews': 4.0, 'Mick LaSalle': 4.0, 'Claudia Puig': 3.5, 'Lisa Rose': 3.5, 'Toby': 4.5,
  32. 'Gene Seymour': 3.5},
  33. 'Just My Luck': {'Larry': 3.5, 'Claudia Puig': 3.0, 'Gene Seymour': 1.5, 'Mick LaSalle': 2.0,
  34. 'Michelle Nichols': 1.0},
  35. 'Superman Returns': {'Gary Coleman': 1.5, 'Jack Matthews': 5.0, 'Mick LaSalle': 3.0, 'Claudia Puig': 4.0,
  36. 'Lisa Rose': 3.5, 'Michael Phillips': 3.5},
  37. 'The Night Listener': {'Larry': 3.5, 'Jack Matthews': 3.0, 'Claudia Puig': 4.5, 'Lisa Rose': 3.0,
  38. 'Gene Seymour': 3.0, 'Michael Phillips': 4.0, 'Michelle Nichols': 4.5},
  39. 'You, Me and Dupree': {'Gary Coleman': 2.0, 'Jack Matthews': 3.5, 'Mick LaSalle': 2.0, 'Claudia Puig': 2.5,
  40. 'Lisa Rose': 2.5, 'Gene Seymour': 3.5, 'Michelle Nichols': 3.5},
  41. 'Catch Me If You Can': {'Lisa Rose': 3.0, 'Michael Phillips': 2.5, 'Jack Matthews': 4.5, 'Gary Coleman': 1.5},
  42. 'Snitch': {'Larry': 1.5, 'Toby': 5.0, 'Michael Phillips': 2.0, 'Lisa Rose': 3.0, 'Jack Matthews': 4.5}
  43. }
  44.  
  45. def sim_distance(oceni, person1, person2):
  46. # Se pravi lista na zaednicki predmeti (filmovi)
  47.  
  48. filmovi1=set(oceni[person1].keys())
  49. filmovi2=set(oceni[person2].keys())
  50. zaednicki = filmovi1.intersection(filmovi2)
  51. # print(filmovi1)
  52. # print(filmovi2)
  53. # print(zaednicki)
  54. # for item in oceni[person1].keys():
  55. # if item in oceni[person2]:
  56. # zaednicki.add(item)
  57. # # ako nemaat zaednicki rejtinzi, vrati 0
  58. if len(zaednicki) == 0: return 0
  59. # # Soberi gi kvadratite na zaednickite razliki
  60. suma = 0.0
  61. for item in zaednicki:
  62. ocena1 = oceni[person1][item]
  63. ocena2 = oceni[person2][item]
  64. suma += (ocena1 - ocena2) ** 2
  65. # print(item, person1, ocena1, person2, ocena2)
  66.  
  67. return round(1 / (1 + sqrt(suma)),2)
  68.  
  69.  
  70.  
  71.  
  72. # In[12]:
  73.  
  74.  
  75.  
  76. # Go vrakja koeficientot na Pearsonova korelacija pomegu p1 i p2 (licnost 1 i licnost 2)
  77. # Vrednostite se pomegu -1 i 1
  78. def sim_pearson(oceni, p1, p2):
  79. # Se kreira recnik vo koj ke se cuvaat predmetite (filmovi) koi se oceneti od dvajcata
  80. # Vo recnikot ni se vazni samo klucevite za da gi cuvame iminjata na filmovite koi se zaednicki, a vrednostite ne ni se vazni
  81. zaednicki = set()
  82. for item in oceni[p1]:
  83. if item in oceni[p2]:
  84. zaednicki.add(item)
  85.  
  86. # Se presmetuva brojot na predmeti oceneti od dvajcata
  87. n = len(zaednicki)
  88.  
  89. # Ako nemaat zaednicki predmeti vrati korelacija 0
  90. if n == 0: return 0
  91.  
  92. # Soberi gi zaednickite oceni (rejtinzi) za sekoja licnost posebno
  93. sum1 = 0
  94. sum2 = 0
  95.  
  96. # Soberi gi kvadratite od zaednickite oceni (rejtinzi) za sekoja licnost posebno
  97. sum1Sq = 0
  98. sum2Sq = 0
  99.  
  100. # Soberi gi proizvodite od ocenite na dvete licnosti
  101. pSum = 0
  102. for item in zaednicki:
  103. ocena1 = oceni[p1][item]
  104. ocena2 = oceni[p2][item]
  105. sum1 += ocena1
  106. sum1Sq += ocena1 ** 2
  107. sum2 += ocena2
  108. sum2Sq += ocena2 ** 2
  109. pSum += ocena1 * ocena2
  110.  
  111. # Presmetaj go koeficientot na korelacija
  112. num = pSum - (sum1 * sum2 / n)
  113. den = sqrt((sum1Sq - pow(sum1, 2) / n) * (sum2Sq - pow(sum2, 2) / n))
  114. if den == 0: return 0
  115. r = num / den
  116. return round(r,2)
  117.  
  118.  
  119.  
  120.  
  121. def topMatches_inline(oceni, person, n=5, similarity=sim_pearson):
  122. scores = [(similarity(oceni, person, other), other)
  123. for other in oceni if other != person]
  124. # Se sortira listata vo rastecki redosled
  125. scores.sort()
  126. # Se prevrtuva za najslicnite (so najgolema vrednost) da bidat prvi
  127. scores.reverse()
  128. return scores[0:n]
  129.  
  130.  
  131. def topMatches(oceni, person, n=5, similarity=sim_pearson):
  132. scores = []
  133. for person2 in oceni.keys():
  134. if person != person2:
  135. s = similarity(oceni, person, person2)
  136. scores.append((s, person2))
  137. # Se sortira listata vo rastecki redosled
  138. scores.sort()
  139. # Se prevrtuva za najslicnite (so najgolema vrednost) da bidat prvi
  140. scores.reverse()
  141. if n is None:
  142. return scores
  143. else:
  144. return scores[0:n]
  145.  
  146.  
  147.  
  148.  
  149.  
  150.  
  151.  
  152.  
  153.  
  154.  
  155.  
  156. # In[52]:
  157.  
  158.  
  159. def transformoceni(oceni):
  160. result = {}
  161. for person in oceni.keys():
  162. for item in oceni[person]:
  163. result.setdefault(item, {})
  164. # Zameni gi mestata na licnosta i predmetot
  165. result[item][person] = oceni[person][item]
  166. return result
  167.  
  168. def getRecommendations(oceni, person, similarity=sim_pearson, min_zaednicki=None):
  169. totals = {}
  170. simSums = {}
  171. for person2 in oceni.keys():
  172. # Za da ne se sporeduva so samiot sebe
  173. if person2 == person: continue
  174. filmovi1=set(oceni[person].keys())
  175. filmovi2=set(oceni[person2].keys())
  176. zaednicki = filmovi1.intersection(filmovi2)
  177. # ova e ako se bara minimum zaednicki filmovi
  178. # za da se zemat vo predvid ocenite na drugiot korisnik
  179. if min_zaednicki and len(zaednicki)<min_zaednicki:
  180. #print('So korisnikot', person2, 'imame samo',len(zaednicki),'filmovi, pa go preskoknuvame')
  181. continue
  182. sim = similarity(oceni, person, person2)
  183. # print(person,person2,sim)
  184. # ne se zemaat vo predvid rezultati <= 0
  185. if sim <= 0: continue
  186. #print(person,person2,sim)
  187. for item in oceni[person2].keys():
  188. # print(item, oceni[person].get(item, None), oceni[person2].get(item, None))
  189. # za tekovniot korisnik gi zemame samo filmovite sto gi nemame veke gledano
  190. if item not in oceni[person]: # or oceni[person][item] == 0:
  191. # similarity * Score (Slicnost * Ocena)
  192. #print(item, sim, oceni[person2][item], sim* oceni[person2][item])
  193. totals.setdefault(item, 0)
  194. totals[item] += oceni[person2][item] * sim
  195.  
  196. # Sumuma na slicnosti
  197. simSums.setdefault(item, 0)
  198. simSums[item] += sim
  199. #print()
  200. # return
  201. #print()
  202. # Kreiranje na normalizirana lista so rejtinzi
  203. # rankings = [(total / simSums[item], item) for item, total in totals.items()]
  204. rankings = []
  205. for item, weighted_score in totals.items():
  206. sim_total = simSums[item]
  207. my_score = round(weighted_score / sim_total, 1)
  208. #print(item, weighted_score, sim_total, my_score)
  209. rankings.append((my_score, item))
  210.  
  211. # Sortiranje na listata vo rastecki redosled
  212. rankings.sort(reverse=True)
  213. # Prevrtuvanje na lista za najgolemite vrednosti da bidat napred
  214. # rankings.reverse()
  215. return rankings
  216. # In[77]:
  217.  
  218.  
  219. def item_based(critics, person1, n=3):
  220. oceni_po_film = transformoceni(critics)
  221. similarity_per_item = {}
  222. for item in critics[person1].keys():
  223. similar_items = topMatches(oceni_po_film, item, n=None)
  224. my_rating = critics[person1][item]
  225.  
  226. for similarity, item2 in similar_items:
  227. if item2 in critics[person1] or similarity <= 0:
  228. # print('Slicnost', similarity, 'na', item,'so', item2)
  229. continue
  230. weight= similarity * my_rating
  231. # print('Slicnost', similarity, 'na', item,'so', item2, weight)
  232. similarity_per_item.setdefault(item2, [])
  233. similarity_per_item[item2].append(weight)
  234. # print(item, my_rating, list(similarity_per_item.items()))
  235. similarity_per_item_avg = []
  236. import numpy as np
  237. for item in similarity_per_item:
  238. #print(item, similarity_per_item[item])
  239. avg_sim = np.mean(similarity_per_item[item])
  240. similarity_per_item_avg.append((avg_sim, item))
  241. similarity_per_item_avg.sort(reverse=True)
  242. return similarity_per_item_avg[:n]
  243.  
  244.  
  245.  
  246.  
  247.  
  248.  
  249. if __name__ == "__main__":
  250. korisnik = 'Gary Coleman'
  251.  
  252. preporaki = getRecommendations(oceniPoKorisnici, korisnik)
  253. preporaki = preporaki[:3]
  254. rezultati = [ime for brojka, ime in preporaki]
  255. rezultati.sort()
  256. print('user-based :', rezultati)
  257.  
  258. preporaki2 = item_based(oceniPoKorisnici, korisnik)
  259. preporaki2 = preporaki2[:3]
  260. rezultati2 = [ime for brojka, ime in preporaki2]
  261. rezultati2.sort()
  262. print('item-based :', rezultati2)
Add Comment
Please, Sign In to add comment
Public Pastes
We use cookies for various purposes including analytics. By continuing to use Pastebin, you agree to our use of cookies as described in the Cookies Policy.  OK, I Understand
Not a member of Pastebin yet?
Sign Up, it unlocks many cool features!