merchants_embed_store.py 9.3 KB

123456789101112131415161718192021222324252627282930313233343536373839404142434445464748495051525354555657585960616263646566676869707172737475767778798081828384858687888990919293949596979899100101102103104105106107108109110111112113114115116117118119120121122123124125126127128129130131132133134135136137138139140141142143144145146147148149150151152153154155156157158159160161162163164165166167168169170171172173174175176177178179180181182183184185186187188189190191192193194195196197198199200201202203204205206207208209210211212213214
  1. """
  2. 行业 / 商户语义库-固定结构:
  3. {
  4. "merchant_industry_id": "",
  5. "industry_name": "",
  6. "typical_merchants": [],
  7. "default_expense_nature": "",
  8. "personal_consume_probability": 0.0,
  9. "enterprise_legit_scenarios": [],
  10. "risk_notes": [],
  11. "embedding_text": ""
  12. }
  13. """
  14. merchants_embed = [
  15. {
  16. "merchant_industry_id": "IND-EDU-001",
  17. "industry_name": "教育培训与学费",
  18. "typical_merchants": ["培训学校", "教育咨询公司", "语言培训中心"],
  19. "default_expense_nature": "个人消费",
  20. "personal_consume_probability": 0.85,
  21. "enterprise_legit_scenarios": [
  22. "企业统一组织员工培训",
  23. "岗位技能提升培训",
  24. "取得培训协议与培训名单"
  25. ],
  26. "risk_notes": [
  27. "单人学费",
  28. "兴趣类培训"
  29. ],
  30. "embedding_text": "教育培训、学费、课程费用通常属于个人消费支出,如语言培训、兴趣课程、个人技能提升等。但在企业统一组织员工培训、岗位能力提升、有培训协议和人员名单的情况下,可作为企业经营性支出。"
  31. }
  32. , {
  33. "merchant_industry_id": "IND-HOBBY-001",
  34. "industry_name": "兴趣班与素质教育",
  35. "typical_merchants": ["艺术培训中心", "音乐学校", "绘画书法机构"],
  36. "default_expense_nature": "个人消费",
  37. "personal_consume_probability": 0.9,
  38. "enterprise_legit_scenarios": [
  39. "企业品牌活动",
  40. "公益合作项目"
  41. ],
  42. "risk_notes": [
  43. "少量学员",
  44. "与主营业务无关"
  45. ],
  46. "embedding_text": "兴趣班、艺术培训、音乐、美术、舞蹈等素质教育类费用高度偏向个人消费,通常不属于企业正常经营支出,除非用于企业品牌推广或公益合作项目。"
  47. }
  48. , {
  49. "merchant_industry_id": "IND-CLOTH-001",
  50. "industry_name": "服饰与鞋帽",
  51. "typical_merchants": ["服装店", "服饰有限公司", "品牌专卖店"],
  52. "default_expense_nature": "个人消费",
  53. "personal_consume_probability": 0.8,
  54. "enterprise_legit_scenarios": [
  55. "企业工装",
  56. "统一制服",
  57. "演出或活动服装"
  58. ],
  59. "risk_notes": [
  60. "品牌服装",
  61. "非统一款式"
  62. ],
  63. "embedding_text": "服饰、鞋帽类支出通常属于个人消费,如日常穿着、品牌服装等。但在企业统一采购工装、制服或用于特定活动演出服装时,可作为企业经营性支出。"
  64. }
  65. , {
  66. "merchant_industry_id": "IND-COSMETIC-001",
  67. "industry_name": "化妆品与护肤品",
  68. "typical_merchants": ["化妆品公司", "美妆专柜"],
  69. "default_expense_nature": "个人消费",
  70. "personal_consume_probability": 0.95,
  71. "enterprise_legit_scenarios": [
  72. "影视拍摄化妆",
  73. "舞台演出妆造"
  74. ],
  75. "risk_notes": [
  76. "高频购买",
  77. "个人品牌商品"
  78. ],
  79. "embedding_text": "化妆品、护肤品、美妆用品高度偏向个人消费支出,通常不作为企业经营成本,仅在影视拍摄、演出化妆等特定业务场景下可能合理。"
  80. }
  81. , {
  82. "merchant_industry_id": "IND-JEWEL-001",
  83. "industry_name": "首饰与奢侈品",
  84. "typical_merchants": ["珠宝店", "奢侈品专卖店"],
  85. "default_expense_nature": "个人消费",
  86. "personal_consume_probability": 0.98,
  87. "enterprise_legit_scenarios": [
  88. "客户礼品",
  89. "商务赠品(需登记)"
  90. ],
  91. "risk_notes": [
  92. "高金额",
  93. "无赠送记录"
  94. ],
  95. "embedding_text": "首饰、珠宝、奢侈品类支出几乎全部属于个人消费,仅在作为客户礼品、商务赠品且有明确登记和用途说明时,才可能视为企业相关支出。"
  96. }
  97. , {
  98. "merchant_industry_id": "IND-MEDICAL-001",
  99. "industry_name": "体检与医疗服务",
  100. "typical_merchants": ["体检中心", "医院"],
  101. "default_expense_nature": "个人消费",
  102. "personal_consume_probability": 0.7,
  103. "enterprise_legit_scenarios": [
  104. "员工年度体检",
  105. "职业健康检查"
  106. ],
  107. "risk_notes": [
  108. "个人名义体检",
  109. "非全员体检"
  110. ],
  111. "embedding_text": "体检费和医疗服务通常为个人消费支出,但在企业统一组织员工体检、职业健康检查等情况下,可作为企业福利或合规经营支出。"
  112. }
  113. , {
  114. "merchant_industry_id": "IND-HEALTH-001",
  115. "industry_name": "保健品",
  116. "typical_merchants": ["保健品公司", "健康产品商店"],
  117. "default_expense_nature": "个人消费",
  118. "personal_consume_probability": 0.9,
  119. "enterprise_legit_scenarios": [
  120. "员工福利发放"
  121. ],
  122. "risk_notes": [
  123. "高频小额",
  124. "个人服用属性明显"
  125. ],
  126. "embedding_text": "保健品、营养品类费用明显偏向个人消费,仅在企业作为员工福利统一采购并发放的情况下,才可能与企业相关。"
  127. }
  128. , {
  129. "merchant_industry_id": "IND-ENT-001",
  130. "industry_name": "电影与娱乐消费",
  131. "typical_merchants": ["电影院", "娱乐票务平台"],
  132. "default_expense_nature": "个人消费",
  133. "personal_consume_probability": 0.85,
  134. "enterprise_legit_scenarios": [
  135. "员工团建",
  136. "企业文化活动"
  137. ],
  138. "risk_notes": [
  139. "单人消费",
  140. "非集体活动"
  141. ],
  142. "embedding_text": "电影票及一般娱乐消费多为个人消费,仅在企业统一组织员工观影、团建或文化活动时,才可能视为企业福利支出。"
  143. }
  144. , {
  145. "merchant_industry_id": "IND-KTV-001",
  146. "industry_name": "KTV与夜间娱乐",
  147. "typical_merchants": ["KTV会所", "娱乐会馆"],
  148. "default_expense_nature": "个人消费",
  149. "personal_consume_probability": 0.9,
  150. "enterprise_legit_scenarios": [
  151. "员工团建",
  152. "客户接待(需说明)"
  153. ],
  154. "risk_notes": [
  155. "夜间高消费",
  156. "无接待对象说明"
  157. ],
  158. "embedding_text": "KTV及夜间娱乐消费通常属于个人消费,仅在明确的员工团建或客户接待场景下,并有相关说明时,才可能与企业经营相关。"
  159. }
  160. , {
  161. "merchant_industry_id": "IND-FIT-001",
  162. "industry_name": "健身与运动服务",
  163. "typical_merchants": ["健身房", "运动中心"],
  164. "default_expense_nature": "个人消费",
  165. "personal_consume_probability": 0.88,
  166. "enterprise_legit_scenarios": [
  167. "员工福利健身卡",
  168. "企业健康计划"
  169. ],
  170. "risk_notes": [
  171. "个人会员卡",
  172. "长期使用"
  173. ],
  174. "embedding_text": "健身卡、运动服务费用高度偏向个人消费,但在企业统一为员工购买健身卡、实施健康福利计划的情况下,可作为企业福利支出。"
  175. }
  176. , {
  177. "merchant_industry_id": "IND-TRAVEL-001",
  178. "industry_name": "旅游与景区消费",
  179. "typical_merchants": ["旅行社", "景区管理公司"],
  180. "default_expense_nature": "个人消费",
  181. "personal_consume_probability": 0.92,
  182. "enterprise_legit_scenarios": [
  183. "员工团建旅游",
  184. "企业会议出行"
  185. ],
  186. "risk_notes": [
  187. "无会议说明",
  188. "自由行"
  189. ],
  190. "embedding_text": "旅游费用、景区门票通常为个人消费,仅在企业统一组织员工团建、会议或培训出行,并有行程与人员说明的情况下,才可能作为企业相关支出。"
  191. }
  192. ]
  193. merchants_embed_update = [
  194. {
  195. "merchant_industry_id": "IND-SHOP-001",
  196. "industry_name": "仓储式零售会员店",
  197. "typical_merchants": ["山姆会员商店", "Costco 开市客", "麦德龙"],
  198. "default_expense_nature": "mixed",
  199. "personal_consume_probability": 0.78,
  200. "enterprise_legit_scenarios": [
  201. "行政物资集中采购",
  202. "节日福利发放",
  203. "会议活动后勤保障"
  204. ],
  205. "risk_notes": [
  206. "个人名义零散消费易混入公款报销",
  207. "需重点核查是否附带采购清单及用途说明",
  208. "高频小额交易可能存在拆单规避审查风险"
  209. ],
  210. "embedding_text": "以山姆会员商店为代表的仓储式零售会员店主要服务于个人会员与企业客户两类群体。其销售商品涵盖食品、日化、家电等多个品类,兼具个人消费与企业采购功能。然而,由于入场需会员资格且多数交易由个人完成,来自此类商户的发票具有较高个人消费概率。企业在进行费用审核时,应对该类商户来源的票据加强用途验证,尤其是单笔金额较小、商品为生活必需品的情形。同时,该类商户也支持企业批量采购,若能提供订单、签收记录或福利发放名单,则可认定为合规支出。"
  211. }
  212. ]