2026年程序設(shè)計(jì)面試題目與解答示例集一、編程實(shí)現(xiàn)題（共3題，每題20分）1.題目（20分）：背景：某電商平臺(tái)需要對(duì)用戶行為數(shù)據(jù)進(jìn)行實(shí)時(shí)統(tǒng)計(jì)，要求統(tǒng)計(jì)當(dāng)前在線用戶中，活躍度最高的前K個(gè)用戶（活躍度定義為：用戶在過(guò)去1小時(shí)內(nèi)產(chǎn)生的請(qǐng)求次數(shù)）。請(qǐng)?jiān)O(shè)計(jì)一個(gè)算法，支持動(dòng)態(tài)添加用戶請(qǐng)求，并實(shí)時(shí)返回前K活躍用戶。要求：-使用Python實(shí)現(xiàn)，數(shù)據(jù)結(jié)構(gòu)需考慮時(shí)間復(fù)雜度和空間復(fù)雜度。-示例輸入：-添加用戶請(qǐng)求序列：`["user1","user2","user1","user3","user2","user1"]`-查詢前K活躍用戶（K=2）：應(yīng)返回`["user1","user2"]`（假設(shè)時(shí)間戳隱含在添加順序中）。2.題目（20分）：背景：在金融風(fēng)控場(chǎng)景中，需要檢測(cè)交易流水中的異常金額（例如，單筆金額超過(guò)閾值的5倍）。給定一個(gè)包含時(shí)間戳和金額的交易列表，要求實(shí)現(xiàn)一個(gè)函數(shù)，返回所有可疑交易（金額異常或交易過(guò)于頻繁）。要求：-使用Java實(shí)現(xiàn)，需考慮高并發(fā)場(chǎng)景下的線程安全。-示例輸入：javaList<Transaction>transactions=Arrays.asList(newTransaction("tx1",1000,System.currentTimeMillis()-1000),newTransaction("tx2",5000,System.currentTimeMillis()-500),newTransaction("tx3",2000,System.currentTimeMillis()-2000),newTransaction("tx1",3000,System.currentTimeMillis()));-設(shè)閾值為1000，異常條件包括：金額>5000或同一用戶2秒內(nèi)交易超過(guò)2次。3.題目（20分）：背景：某物流公司需要優(yōu)化配送路線，給定多個(gè)配送點(diǎn)（坐標(biāo)）和配送任務(wù)（起點(diǎn)-終點(diǎn)），要求計(jì)算最短的總配送路徑?？墒褂脠D算法實(shí)現(xiàn)，需考慮動(dòng)態(tài)添加任務(wù)的情況。要求：-使用C++實(shí)現(xiàn)，支持動(dòng)態(tài)更新任務(wù)隊(duì)列。-示例輸入：cppvector<pair<string,pair<int,int>>>points={{"A",{0,0}},{"B",{1,2}},{"C",{3,1}}};vector<pair<string,string>>tasks={{"A","B"},{"B","C"}};-計(jì)算最短路徑（可假設(shè)距離函數(shù)為曼哈頓距離）。二、算法設(shè)計(jì)題（共2題，每題25分）1.題目（25分）：背景：在社交網(wǎng)絡(luò)中，需要檢測(cè)用戶之間的潛在欺詐關(guān)系（例如，通過(guò)共同好友過(guò)多可能為虛假賬戶）。給定用戶好友關(guān)系圖，要求設(shè)計(jì)算法識(shí)別高關(guān)聯(lián)度的用戶對(duì)（例如，共同好友數(shù)超過(guò)閾值）。要求：-使用DFS/BFS實(shí)現(xiàn)，需考慮圖數(shù)據(jù)結(jié)構(gòu)的存儲(chǔ)。-示例輸入：pythongraph={"user1":["user2","user3"],"user2":["user1","user4"],"user3":["user1"],"user4":["user2"]}-設(shè)閾值為2，應(yīng)識(shí)別出`("user1","user2")`和`("user2","user4")`。2.題目（25分）：背景：在自然語(yǔ)言處理中，需要實(shí)現(xiàn)詞性標(biāo)注（POStagging）的動(dòng)態(tài)規(guī)劃算法。給定句子和詞典，要求標(biāo)注每個(gè)詞的詞性（如名詞、動(dòng)詞）。要求：-使用Python實(shí)現(xiàn)，需考慮狀態(tài)轉(zhuǎn)移方程。-示例輸入：pythonsentence=["我","愛(ài)","編程"]vocab={"我":"PRON","愛(ài)":"VERB","編程":"NOUN"}-輸出標(biāo)注結(jié)果：`[("我","PRON"),("愛(ài)","VERB"),("編程","NOUN")]`。三、系統(tǒng)設(shè)計(jì)題（共1題，30分）1.題目（30分）：背景：設(shè)計(jì)一個(gè)高并發(fā)的短鏈接系統(tǒng)（如tinyURL），需支持URL生成、解析和分布式緩存。要求：-描述系統(tǒng)架構(gòu)（數(shù)據(jù)庫(kù)、緩存、負(fù)載均衡）。-實(shí)現(xiàn)核心的URL編碼/解碼算法（可使用Base62）。-示例輸入：javagenerate("");//輸出"a1b2"parse("a1b2");//返回""答案與解析一、編程實(shí)現(xiàn)題1.答案（Python）：pythonimportheapqclassTopKActiveUsers:def__init__(self,k):self.k=kself.user_counts={}self.heap=[]defadd_request(self,user):ifuserinself.user_counts:self.user_counts[user]+=1else:self.user_counts[user]=1iflen(self.heap)<self.k:heapq.heappush(self.heap,(-self.user_counts[user],user))elifself.user_counts[user]>-self.heap[0][0]:heapq.heapreplace(self.heap,(-self.user_counts[user],user))defget_top_k(self):return[userfor_,userinsorted(self.heap,reverse=True)]示例tk=TopKActiveUsers(2)requests=["user1","user2","user1","user3","user2","user1"]forreqinrequests:tk.add_request(req)print(tk.get_top_k())#輸出["user1","user2"]解析：-使用字典`user_counts`記錄用戶請(qǐng)求次數(shù)，堆`heap`維護(hù)前K大用戶。-時(shí)間復(fù)雜度：`O(logk)`（每次添加操作），空間復(fù)雜度：`O(k)`。2.答案（Java）：javaimportjava.util.;classTransaction{Stringid;intamount;longtimestamp;publicTransaction(Stringid,intamount,longtimestamp){this.id=id;this.amount=amount;this.timestamp=timestamp;}}publicclassFraudDetection{privatestaticfinalintTHRESHOLD=5000;privateMap<String,Transaction>recentTransactions=newConcurrentHashMap<>();privateMap<String,Integer>userFrequency=newConcurrentHashMap<>();publicvoidaddTransaction(Transactiontx){if(tx.amount>THRESHOLD){System.out.println("Suspicious:"+tx);}longcurrentTime=System.currentTimeMillis();recentTransactions.values().removeIf(t->currentTime-t.timestamp>2000);recentTransactions.put(tx.id,tx);userFrequency.put(tx.id,userFrequency.getOrDefault(tx.id,0)+1);if(userFrequency.get(tx.id)>2){System.out.println("Suspicious:"+tx);}}publicstaticvoidmain(String[]args){FraudDetectionfd=newFraudDetection();List<Transaction>txs=Arrays.asList(newTransaction("tx1",1000,System.currentTimeMillis()-1000),newTransaction("tx2",5000,System.currentTimeMillis()-500),newTransaction("tx3",2000,System.currentTimeMillis()-2000),newTransaction("tx1",3000,System.currentTimeMillis()));for(Transactiontx:txs){fd.addTransaction(tx);}}}解析：-使用`ConcurrentHashMap`確保線程安全。-檢測(cè)條件：金額異?；蛴脩纛l率過(guò)高。3.答案（C++）：cppinclude<vector>include<string>include<unordered_map>include<queue>include<algorithm>usingnamespacestd;structPoint{stringid;intx,y;};structEdge{intfrom,to;intdist;};intmanhattan(pair<int,int>a,pair<int,int>b){returnabs(a.first-b.first)+abs(a.second-b.second);}vector<int>dijkstra(constvector<vector<Edge>>&graph,intstart,intn){vector<int>dist(n,INT_MAX);priority_queue<pair<int,int>,vector<pair<int,int>>,greater<pair<int,int>>>pq;dist[start]=0;pq.push({0,start});while(!pq.empty()){intd=pq.top().first;intu=pq.top().second;pq.pop();if(d>dist[u])continue;for(auto&e:graph[u]){intv=e.to;intalt=d+e.dist;if(alt<dist[v]){dist[v]=alt;pq.push({alt,v});}}}returndist;}intmain(){vector<Point>points={{"A",{0,0}},{"B",{1,2}},{"C",{3,1}}};vector<vector<Edge>>graph(3,vector<Edge>());graph[0].push_back({0,1,manhattan({0,0},{1,2})});graph[1].push_back({1,0,manhattan({1,2},{0,0})});graph[1].push_back({1,2,manhattan({1,2},{3,1})});graph[2].push_back({2,1,manhattan({3,1},{1,2})});vector<int>dist=dijkstra(graph,0,3);for(intd:dist)cout<<d<<"";return0;}解析：-使用Dijkstra算法計(jì)算最短路徑，動(dòng)態(tài)更新任務(wù)隊(duì)列可通過(guò)維護(hù)優(yōu)先隊(duì)列實(shí)現(xiàn)。二、算法設(shè)計(jì)題1.答案（Python）：pythondeffind_highly_connected_pairs(graph,threshold):common_friends={}foruser,friendsingraph.items():forfriendinfriends:iffriendnotingraph:continueshared=set(friends).intersection(graph[friend])iflen(shared)>=threshold:if(user,friend)notincommon_friends:common_friends[(user,friend)]=len(shared)elif(friend,user)notincommon_friends:common_friends[(friend,user)]=len(shared)return[pairforpair,countincommon_friends.items()ifcount>=threshold]示例graph={"user1":["user2","user3"],"user2":["user1","user4"],"user3":["user1"],"user4":["user2"]}print(find_highly_connected_pairs(graph,2))#輸出[("user1","user2"),("user2","user4")]解析：-遍歷圖，統(tǒng)計(jì)每對(duì)用戶的共同好友數(shù)量，超過(guò)閾值則記錄。2.答案（Python）：pythondefpos_tagging(sentence,vocab):n=len(sentence)dp=[[""]nfor_inrange(2)]foriinrange(n):word=sentence[i]ifwordinvocab:forjinrange(i+1):ifj==0:dp[i%2][j]=(word,vocab[word])else:dp[i%2][j]=max(dp[(i-1)%2][j-1],dp[(i-1)%2][j],key=lambdax:x[1]ifx[1]else-1)[0]dp[i%2][i]=(word,vocab[word])returndp[(n-1)%2]示例sentence=["我","愛(ài)","編程"]vocab={"我":"PRON","愛(ài)":"VERB","編程":"NOUN"}print(pos_tagging(sentence,vocab))#輸出[("我","PRON"),("愛(ài)","VERB"),("編程","NOUN")]解析：-動(dòng)態(tài)規(guī)劃，狀態(tài)轉(zhuǎn)移基于前一個(gè)詞的詞性。三、系統(tǒng)設(shè)計(jì)題答案（Java）：javaimportjava.util.;publicclassShortLinkSystem{privatestaticfinalStringBASE62="0123456789abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ";privateMap<String,String>urlMap=newConcurrentHashMap<>();privateMap<String,String>reverseMap=newConcurrentHashMap<>();privateintcount=0;publicString