拆包

@@ -4,7 +4,7 @@ import com.yoho.search.base.utils.SearchCollectionUtils;
 import com.yoho.search.base.utils.Transfer;
 import com.yoho.search.core.personalized.models.PersonalizedSearch;
 import com.yoho.search.core.personalized.models.SortPriceAreas;
-import com.yoho.search.service.recall.beans.persional.ProductFeatureFactorComponent;
+import com.yoho.search.service.recall.helper.W2vFeatureCalculator;
 import com.yoho.search.service.recall.config.RecallConfigConstants;
 import com.yoho.search.service.recall.config.RecallConfigService;
 import com.yoho.search.service.recall.models.personal.UserFeatureFactor;
@@ -34,8 +34,6 @@ public class UserRecallResponseBuilder {
     @Autowired
     private PersonalVectorFeatureSearch personalVectorFeatureSearch;
     @Autowired
-    private ProductFeatureFactorComponent productFeatureFactorComponent;
-    @Autowired
     private SknBaseInfoBaseService sknBaseInfoBaseService;
     @Autowired
     private RecallConfigService recallConfigService;
@@ -200,7 +198,7 @@ public class UserRecallResponseBuilder {
                 continue;
             }
             //3)向量计算-如果为1，则使用百分之1的人气替代
-            score = productFeatureFactorComponent.calProductFeatureFactor(userFeatureFactor, sknResult.getFactor());
+            score = W2vFeatureCalculator.calProductFeatureFactor(userFeatureFactor, sknResult.getFactor());
             if (score == 0) {
                 score = sknResult.getHeatValue() / 100;
             }

-package com.yoho.search.service.recall.beans.persional;
+package com.yoho.search.service.recall.helper;
 
 import com.yoho.search.base.helper.Word2VectorCalculator;
 import com.yoho.search.base.utils.ConvertUtils;
@@ -8,19 +8,17 @@ import org.apache.commons.collections.CollectionUtils;
 import org.apache.commons.lang.StringUtils;
 import org.slf4j.Logger;
 import org.slf4j.LoggerFactory;
-import org.springframework.stereotype.Component;
 
 import java.util.List;
 
-@Component
-public class ProductFeatureFactorComponent {
+public class W2vFeatureCalculator {
 
-    private static final Logger logger = LoggerFactory.getLogger(ProductFeatureFactorComponent.class);
+    private static final Logger logger = LoggerFactory.getLogger(W2vFeatureCalculator.class);
 
     private static double baseConstant = 1;
     private static double factorConstant = 1;
 
-    public double calProductFeatureFactor(UserFeatureFactor userFeatureFactor, String productFeatureFactor) {
+    public static double calProductFeatureFactor(UserFeatureFactor userFeatureFactor, String productFeatureFactor) {
         try {
             //商品向量不存在，则返回
             if (StringUtils.isBlank(productFeatureFactor)) {
@@ -57,7 +55,7 @@ public class ProductFeatureFactorComponent {
         PersonalizedSearch personalizedSearch = new PersonalizedSearch("1", "20180408", array);
         UserFeatureFactor userFeatureFactor = new UserFeatureFactor(personalizedSearch);
         String productFeatureFactor = "20180408|1.342045,-0.547933,0.291732,-0.056515,-0.182701,0.31113,0.151578,0.087678,-0.045536,-0.525699,-0.394715,-0.103153,-0.05575,-0.540641,0.028046,-0.193109,-0.003591,0.180923,0.290261,0.532309,-0.202463,-0.047271,-0.246197,0.324561,0.188814,0.36475,0.079007,0.455753,-0.11848,-0.135874,-0.187155,-0.055342,-0.12525,0.210669,-0.388331,-0.197123,0.132309,-0.4231,0.217752,-0.203266,0.190836,0.373428,-0.0102,-0.038654,0.2379,0.044424,0.071826,-0.201054,0.257434,0.141901,-0.390064,0.437099,0.559701,-0.040162,-0.193089,0.442338,-0.141678,-0.049696,0.315545,-0.028972,0.278694,-0.064345,-0.327943,0.103025,-0.40344,-0.34269,-0.237931,0.287046,0.139693,-0.38454,0.019959,-0.156907,0.374996,-0.074558,-0.019391,0.050522,0.315171,0.211605,-0.15418,0.502362,0.10184,0.153274,0.592659,-0.010284,0.28029,0.319741,-0.164559,0.286884,0.420483,-0.628866,-0.172259,0.027954,-0.411674,0.376585,0.322832,0.352039,0.078705,0.045152,0.139083,-0.164182";
-        System.out.println(new ProductFeatureFactorComponent().calProductFeatureFactor(userFeatureFactor, productFeatureFactor));
+        System.out.println(calProductFeatureFactor(userFeatureFactor, productFeatureFactor));
     }
 
 }

@@ -6,14 +6,14 @@ import com.yoho.search.cache.beans.AbstractCacheBean;
 import com.yoho.search.core.personalized.models.PersonalizedSearch;
 import com.yoho.search.core.personalized.models.SortBrand;
 import com.yoho.search.core.personalized.models.UserPersonalFactorRsp;
-import com.yoho.search.service.recall.beans.persional.ProductFeatureFactorComponent;
 import com.yoho.search.service.recall.beans.persional.UserPersionalFactorComponent;
-import com.yoho.search.service.recall.strategy.IStrategy;
 import com.yoho.search.service.recall.beans.vector.BrandVectorCacheBean;
+import com.yoho.search.service.recall.helper.W2vFeatureCalculator;
 import com.yoho.search.service.recall.models.common.ParamQueryFilter;
 import com.yoho.search.service.recall.models.personal.BrandVectorScore;
 import com.yoho.search.service.recall.models.personal.UserFeatureFactor;
 import com.yoho.search.service.recall.models.req.RecallRequest;
+import com.yoho.search.service.recall.strategy.IStrategy;
 import com.yoho.search.service.scorer.personal.PersonalVectorFeatureSearch;
 import org.apache.commons.collections.MapUtils;
 import org.springframework.beans.factory.annotation.Autowired;
@@ -36,8 +36,6 @@ public class AggBrandProductCacheBean extends AbstractCacheBean<BrandProductRequ
     private BatchBrandProductCacheBean batchBrandProductCacheBean;
     @Autowired
     private PersonalVectorFeatureSearch personalVectorFeatureSearch;
-    @Autowired
-    private ProductFeatureFactorComponent productFeatureFactorComponent;
 
     /**
      * 入口
@@ -115,7 +113,7 @@ public class AggBrandProductCacheBean extends AbstractCacheBean<BrandProductRequ
 
     private void doCalScoreAndSort(List<BrandProduct> productList, UserFeatureFactor userFeatureFactor) {
         for (BrandProduct product : productList) {
-            double score = productFeatureFactorComponent.calProductFeatureFactor(userFeatureFactor, product.getProductFeatureFactor());
+            double score = W2vFeatureCalculator.calProductFeatureFactor(userFeatureFactor, product.getProductFeatureFactor());
             if (score == 0) {
                 score = (double) product.getHeatValue() / 100;
             }