shared-dg04-hi/hi-backend
7
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Merge branch 'main' of https://github.com/dg04-hi/hi-backend

Browse Source
This commit is contained in:
UNGGU0704 2025-06-18 16:12:41 +09:00
commit d86c6f4d74
12 changed files with 299 additions and 16 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

99
analytics/src/main/java/com/ktds/hi/analytics/biz/service/AnalyticsService.java
View File

@ -4,6 +4,7 @@ import com.ktds.hi.analytics.biz.domain.ActionPlan;
import com.ktds.hi.analytics.biz.domain.Analytics;
import com.ktds.hi.analytics.biz.domain.AiFeedback;
import com.ktds.hi.analytics.biz.domain.PlanStatus;
import com.ktds.hi.analytics.biz.domain.SentimentType;
import com.ktds.hi.analytics.biz.usecase.in.AnalyticsUseCase;
import com.ktds.hi.analytics.biz.usecase.out.*;
import com.ktds.hi.analytics.infra.dto.*;
@ -13,10 +14,14 @@ import org.springframework.cache.annotation.Cacheable;
import org.springframework.stereotype.Service;
import org.springframework.transaction.annotation.Transactional;
import java.time.Duration;
import java.time.LocalDate;
import java.time.LocalDateTime;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.Optional;
import java.util.stream.Collectors;
/**
* 분석 서비스 구현 클래스 (수정버전)
@ -35,6 +40,7 @@ public class AnalyticsService implements AnalyticsUseCase {
private final CachePort cachePort;
private final EventPort eventPort;
private final ActionPlanPort actionPlanPort; // 추가된 의존성
@Override
// @Cacheable(value = "storeAnalytics", key = "#storeId")
@ -270,8 +276,10 @@ public class AnalyticsService implements AnalyticsUseCase {
.totalReviews(0)
.positiveReviewCount(0)
.negativeReviewCount(0)
.neutralReviewCount(0)
.positiveRate(0.0)
.negativeRate(0.0)
.neutralRate(0.0)
.analysisDate(LocalDate.now())
.build();
@ -280,8 +288,10 @@ public class AnalyticsService implements AnalyticsUseCase {
}
// 3. 응답 생성
int positiveCount = countPositiveReviews(recentReviews);
int negativeCount = countNegativeReviews(recentReviews);
ReviewSentimentCount sentimentCount = analyzeReviewSentiments(recentReviews);
int positiveCount = sentimentCount.getPositiveCount();
int negativeCount = sentimentCount.getNegativeCount();
int neutralCount = sentimentCount.getNeutralCount();
int totalCount = recentReviews.size();
ReviewAnalysisResponse response = ReviewAnalysisResponse.builder()
@ -289,8 +299,10 @@ public class AnalyticsService implements AnalyticsUseCase {
.totalReviews(totalCount)
.positiveReviewCount(positiveCount)
.negativeReviewCount(negativeCount)
.positiveRate(Math.floor((double) positiveCount / totalCount * 100) / 10.0)
.negativeRate(Math.floor((double) negativeCount / totalCount * 100) / 10.0)
.neutralReviewCount(neutralCount)
.positiveRate(Math.floor((double) positiveCount / totalCount * 1000) / 10.0)
.negativeRate(Math.floor((double) negativeCount / totalCount * 1000) / 10.0)
.neutralRate(Math.floor((double) neutralCount / totalCount * 1000) / 10.0)
.analysisDate(LocalDate.now())
.build();
@ -305,7 +317,75 @@ public class AnalyticsService implements AnalyticsUseCase {
throw new RuntimeException("리뷰 분석에 실패했습니다.", e);
}
}
/**
* 기존 analyzeReviewSentiments 메서드를 대량 분석 방식으로 개선
* 개별 AI 호출 대신 한 번의 호출로 모든 리뷰 분석
*/
private ReviewSentimentCount analyzeReviewSentiments(List<String> reviews) {
log.info("LLM 기반 리뷰 감정 분석 시작: 총 리뷰 수={}", reviews.size());
try {
if (reviews.isEmpty()) {
return new ReviewSentimentCount(0, 0, 0);
}
// 유효한 리뷰만 필터링
List<String> validReviews = reviews.stream()
.filter(review -> review != null && !review.trim().isEmpty())
.collect(Collectors.toList());
if (validReviews.isEmpty()) {
return new ReviewSentimentCount(0, 0, 0);
}
// 기존 개별 분석 대신 대량 분석 사용
Map<SentimentType, Integer> sentimentCounts = aiServicePort.analyzeBulkSentiments(validReviews);
int positiveCount = sentimentCounts.get(SentimentType.POSITIVE);
int negativeCount = sentimentCounts.get(SentimentType.NEGATIVE);
int neutralCount = sentimentCounts.get(SentimentType.NEUTRAL);
ReviewSentimentCount result = new ReviewSentimentCount(positiveCount, negativeCount, neutralCount);
log.info("리뷰 감정 분석 완료: 긍정={}, 부정={}, 중립={}, 전체={}",
positiveCount, negativeCount, neutralCount, validReviews.size());
return result;
} catch (Exception e) {
log.error("리뷰 감정 분석 중 전체 오류 발생, fallback 사용", e);
// 오류 시 기존 가정값 사용
int total = reviews.size();
return new ReviewSentimentCount(
(int) (total * 0.6), // 60% 긍정
(int) (total * 0.2), // 20% 부정
total - (int) (total * 0.6) - (int) (total * 0.2) // 나머지 중립
);
}
}
/**
* 리뷰 감정 분석 결과를 담는 내부 클래스
*/
public static class ReviewSentimentCount {
private final int positiveCount;
private final int negativeCount;
private final int neutralCount;
public ReviewSentimentCount(int positiveCount, int negativeCount, int neutralCount) {
this.positiveCount = positiveCount;
this.negativeCount = negativeCount;
this.neutralCount = neutralCount;
}
public int getPositiveCount() { return positiveCount; }
public int getNegativeCount() { return negativeCount; }
public int getNeutralCount() { return neutralCount; }
public int getTotalCount() { return positiveCount + negativeCount + neutralCount; }
}
// private 메서드들
@Transactional
public Analytics generateNewAnalytics(Long storeId) {
@ -429,15 +509,6 @@ public class AnalyticsService implements AnalyticsUseCase {
return "추천사항이 없습니다.";
}
private int countPositiveReviews(List<String> reviews) {
// 실제로는 AI 서비스를 통한 감정 분석 필요
return (int) (reviews.size() * 0.6); // 60% 가정
}
private int countNegativeReviews(List<String> reviews) {
// 실제로는 AI 서비스를 통한 감정 분석 필요
return (int) (reviews.size() * 0.2); // 20% 가정
}
@Override
@Transactional

10
analytics/src/main/java/com/ktds/hi/analytics/biz/usecase/out/AIServicePort.java
View File

@ -4,6 +4,7 @@ import com.ktds.hi.analytics.biz.domain.AiFeedback;
import com.ktds.hi.analytics.biz.domain.SentimentType;
import java.util.List;
import java.util.Map;
/**
* AI 서비스 포트 인터페이스
@ -20,6 +21,15 @@ public interface AIServicePort {
* 감정 분석
*/
SentimentType analyzeSentiment(String content);
/**
* 대량 리뷰 감정 분석 (새로 추가)
* 여러 리뷰를 한 번에 분석하여 긍정/부정/중립 개수 반환
*
* @param reviews 분석할 리뷰 목록
* @return 감정 타입별 개수 맵
*/
Map<SentimentType, Integer> analyzeBulkSentiments(List<String> reviews);
/**
* 실행 계획 생성

2
analytics/src/main/java/com/ktds/hi/analytics/infra/dto/ReviewAnalysisResponse.java
View File

@ -20,7 +20,9 @@ public class ReviewAnalysisResponse {
private Integer totalReviews;
private Integer positiveReviewCount;
private Integer negativeReviewCount;
private Integer neutralReviewCount;
private Double positiveRate;
private Double negativeRate;
private Double neutralRate;
private LocalDate analysisDate;
}

112
analytics/src/main/java/com/ktds/hi/analytics/infra/gateway/AIServiceAdapter.java
View File

@ -30,8 +30,10 @@ import org.springframework.web.client.RestTemplate;
import java.time.LocalDateTime;
import java.util.ArrayList;
import java.util.Arrays;
import java.util.HashMap;
import java.util.List;
import java.util.Map;
import java.util.stream.Collectors;
/**
* AI 서비스 어댑터 클래스
@ -99,6 +101,116 @@ public class AIServiceAdapter implements AIServicePort {
}
}
@Override
public Map<SentimentType, Integer> analyzeBulkSentiments(List<String> reviews) {
log.info("대량 리뷰 감정 분석 시작: 리뷰 수={}", reviews.size());
try {
if (reviews.isEmpty()) {
return createEmptyResultMap();
}
// 유효한 리뷰만 필터링
List<String> validReviews = reviews.stream()
.filter(review -> review != null && !review.trim().isEmpty())
.collect(Collectors.toList());
if (validReviews.isEmpty()) {
return createEmptyResultMap();
}
// 리뷰를 번호와 함께 포맷팅
StringBuilder reviewsText = new StringBuilder();
for (int i = 0; i < validReviews.size(); i++) {
reviewsText.append(String.format("%d. %s\n", i + 1, validReviews.get(i)));
}
String prompt = String.format(
"""
다음 리뷰들을 분석하여 긍정, 부정, 중립의 개수를 세어주세요.
리뷰 목록:
%s
결과를 다음 JSON 형식으로만 답변해주세요:
{
"positive": 긍정_개수,
"negative": 부정_개수,
"neutral": 중립_개수
}
다른 설명은 하지 말고 JSON만 답변해주세요.
긍정,부정,중립 개수를 모두 더했을때, 총 리뷰수와 동일해야 합니다.
정확하게 세어주세요.
""",
reviewsText.toString()
);
// 기존 callOpenAI 메서드 활용
String result = callOpenAI(prompt);
// 결과 파싱
Map<SentimentType, Integer> sentimentMap = parseBulkSentimentResult(result, validReviews.size());
log.info("대량 리뷰 감정 분석 완료: 긍정={}, 부정={}, 중립={}",
sentimentMap.get(SentimentType.POSITIVE),
sentimentMap.get(SentimentType.NEGATIVE),
sentimentMap.get(SentimentType.NEUTRAL));
return sentimentMap;
} catch (Exception e) {
log.error("대량 리뷰 감정 분석 중 오류 발생, fallback 사용", e);
return createFallbackResultMap(reviews.size());
}
}
private Map<SentimentType, Integer> parseBulkSentimentResult(String result, int totalReviews) {
try {
// 기존 objectMapper 필드 사용
Map<String, Object> jsonResult = objectMapper.readValue(result.trim(), Map.class);
int positive = ((Number) jsonResult.getOrDefault("positive", 0)).intValue();
int negative = ((Number) jsonResult.getOrDefault("negative", 0)).intValue();
int neutral = ((Number) jsonResult.getOrDefault("neutral", 0)).intValue();
// 결과 검증 및 보정
int totalAnalyzed = positive + negative + neutral;
if (totalAnalyzed != totalReviews) {
log.warn("분석 결과 불일치 보정: 분석된 수={}, 실제 리뷰 수={}", totalAnalyzed, totalReviews);
int difference = totalReviews - totalAnalyzed;
neutral += difference;
}
Map<SentimentType, Integer> resultMap = new HashMap<>();
resultMap.put(SentimentType.POSITIVE, Math.max(0, positive));
resultMap.put(SentimentType.NEGATIVE, Math.max(0, negative));
resultMap.put(SentimentType.NEUTRAL, Math.max(0, neutral));
return resultMap;
} catch (Exception e) {
log.error("대량 감정 분석 결과 파싱 실패: {}", result, e);
return createFallbackResultMap(totalReviews);
}
}
private Map<SentimentType, Integer> createEmptyResultMap() {
Map<SentimentType, Integer> result = new HashMap<>();
result.put(SentimentType.POSITIVE, 0);
result.put(SentimentType.NEGATIVE, 0);
result.put(SentimentType.NEUTRAL, 0);
return result;
}
private Map<SentimentType, Integer> createFallbackResultMap(int totalReviews) {
Map<SentimentType, Integer> result = new HashMap<>();
result.put(SentimentType.POSITIVE, (int) (totalReviews * 0.6));
result.put(SentimentType.NEGATIVE, (int) (totalReviews * 0.2));
result.put(SentimentType.NEUTRAL, totalReviews - (int) (totalReviews * 0.6) - (int) (totalReviews * 0.2));
return result;
}
@Override
public SentimentType analyzeSentiment(String content) {

4
analytics/src/main/java/com/ktds/hi/analytics/infra/gateway/EventHubAdapter.java
View File

@ -43,7 +43,7 @@ public class EventHubAdapter {
private final ExecutorService executorService = Executors.newFixedThreadPool(3);
private volatile boolean isRunning = false;
@PostConstruct
// @PostConstruct
public void startEventListening() {
log.info("Event Hub 리스너 시작");
isRunning = true;
@ -52,7 +52,7 @@ public class EventHubAdapter {
executorService.submit(this::listenToReviewEvents);
}
@PreDestroy
// @PreDestroy
public void stopEventListening() {
log.info("Event Hub 리스너 종료");
isRunning = false;

BIN
dump.rdb
View File

Binary file not shown.

BIN
logs/recommend-service.log.2025-06-17.0.gz Normal file
View File

Binary file not shown.

26
store/src/main/java/com/ktds/hi/store/biz/service/StoreService.java
View File

@ -3,6 +3,7 @@ package com.ktds.hi.store.biz.service;
import com.ktds.hi.store.biz.usecase.in.StoreUseCase;
import com.ktds.hi.store.infra.dto.*;
import com.ktds.hi.store.infra.dto.response.StoreListResponse;
import com.ktds.hi.store.infra.gateway.entity.StoreEntity;
import com.ktds.hi.store.infra.gateway.entity.TagEntity;
import com.ktds.hi.store.infra.gateway.repository.StoreJpaRepository;
@ -93,6 +94,27 @@ public class StoreService implements StoreUseCase {
.collect(Collectors.toList());
}
@Override
public List<StoreListResponse> getAllStores() {
List<StoreEntity> stores = storeJpaRepository.findAll();
return stores.stream()
.map(store -> StoreListResponse.builder()
.storeId(store.getId())
.storeName(store.getStoreName())
.address(store.getAddress())
.category(store.getCategory())
.rating(store.getRating())
.reviewCount(store.getReviewCount())
.status("운영중")
.imageUrl(store.getImageUrl())
.operatingHours(store.getOperatingHours())
.build())
.collect(Collectors.toList());
}
@Override
public StoreDetailResponse getStoreDetail(Long storeId) {
@ -157,6 +179,10 @@ public class StoreService implements StoreUseCase {
.build();
}
@Override
public List<StoreSearchResponse> searchStores(String keyword, String category, String tags,
Double latitude, Double longitude, Integer radius,

5
store/src/main/java/com/ktds/hi/store/biz/usecase/in/StoreUseCase.java
View File

@ -1,6 +1,7 @@
package com.ktds.hi.store.biz.usecase.in;
import com.ktds.hi.store.infra.dto.*;
import com.ktds.hi.store.infra.dto.response.StoreListResponse;
import java.util.List;
@ -30,6 +31,8 @@ public interface StoreUseCase {
*/
List<MyStoreListResponse> getMyStores(Long ownerId);
List<StoreListResponse> getAllStores();
/**
* 매장 상세 조회
*
@ -73,4 +76,6 @@ public interface StoreUseCase {
List<StoreSearchResponse> searchStores(String keyword, String category, String tags,
Double latitude, Double longitude, Integer radius,
Integer page, Integer size);
}

14
store/src/main/java/com/ktds/hi/store/infra/controller/StoreController.java
View File

@ -1,11 +1,14 @@
// store/src/main/java/com/ktds/hi/store/infra/controller/StoreController.java
package com.ktds.hi.store.infra.controller;
import com.ktds.hi.store.biz.service.StoreService;
import com.ktds.hi.store.biz.usecase.in.MenuUseCase;
import com.ktds.hi.store.biz.usecase.in.StoreUseCase;
import com.ktds.hi.store.domain.Store;
import com.ktds.hi.store.infra.dto.*;
import com.ktds.hi.common.dto.ApiResponse;
import com.ktds.hi.common.security.JwtTokenProvider;
import com.ktds.hi.store.infra.dto.response.StoreListResponse;
import com.ktds.hi.store.infra.dto.response.StoreMenuListResponse;
import io.swagger.v3.oas.annotations.Operation;
import io.swagger.v3.oas.annotations.Parameter;
@ -38,6 +41,7 @@ import java.util.List;
public class StoreController {
private final StoreUseCase storeUseCase;
private final StoreService storeService;
private final JwtTokenProvider jwtTokenProvider;
private final MenuUseCase menuUseCase;
@ -68,6 +72,14 @@ public class StoreController {
return ResponseEntity.ok(ApiResponse.success(responses, "내 매장 목록 조회 완료"));
}
@GetMapping("/stores/all")
@Operation(summary = "매장 전체 리스트")
public ResponseEntity<ApiResponse<List<StoreListResponse>>> getAllStores() {
List<StoreListResponse> responses = storeUseCase.getAllStores();
return ResponseEntity.ok(ApiResponse.success(responses));
}
@Operation(summary = "매장 상세 조회", description = "매장의 상세 정보를 조회합니다.")
@GetMapping("/{storeId}")
public ResponseEntity<ApiResponse<StoreDetailResponse>> getStoreDetail(
@ -133,6 +145,8 @@ public class StoreController {
@GetMapping("/{storeId}/menus/popular")
@Operation(summary = "매장 인기 메뉴 조회", description = "매장의 인기 메뉴(주문 횟수 기준)를 조회합니다.")
public ResponseEntity<ApiResponse<List<StoreMenuListResponse>>> getPopularMenus(

42
store/src/main/java/com/ktds/hi/store/infra/dto/response/StoreListResponse.java Normal file
View File

@ -0,0 +1,42 @@
package com.ktds.hi.store.infra.dto.response;
import io.swagger.v3.oas.annotations.media.Schema;
import lombok.AllArgsConstructor;
import lombok.Builder;
import lombok.Getter;
import lombok.NoArgsConstructor;
@Getter
@Builder
@NoArgsConstructor
@AllArgsConstructor
@Schema(description = "내 매장 목록 응답")
public class StoreListResponse {
@Schema(description = "매장 ID", example = "1")
private Long storeId;
@Schema(description = "매장명", example = "맛집 한번 가볼래?")
private String storeName;
@Schema(description = "주소", example = "서울시 강남구 테헤란로 123")
private String address;
@Schema(description = "카테고리", example = "한식")
private String category;
@Schema(description = "평점", example = "4.5")
private Double rating;
@Schema(description = "리뷰 수", example = "127")
private Integer reviewCount;
@Schema(description = "운영 상태", example = "운영중")
private String status;
@Schema(description = "운영시간", example = "월-금 09:00-21:00")
private String operatingHours;
@Schema(description = "매장 이미지")
private String imageUrl;
}

1
store/src/main/java/com/ktds/hi/store/infra/gateway/repository/StoreJpaRepository.java
View File

@ -24,6 +24,7 @@ public interface StoreJpaRepository extends JpaRepository<StoreEntity, Long> {
@Query("SELECT s FROM StoreEntity s WHERE s.status = 'ACTIVE' ORDER BY s.rating DESC")
Page<StoreEntity> findAllByOrderByRatingDesc(Pageable pageable);
/**
* 점주 ID로 매장 목록 조회
*/