锦中人工智能助手

import json

from sklearn.feature_extraction.text import TfidfVectorizer

from sklearn.metrics.pairwise import cosine_similarity

# 构建知识库

knowledge_base = {

"questions": [

"如何申请奖学金？",

"怎么查询成绩？",

"学校有哪些社团？",

"如何办理学生证？"

],

"answers": [

"您可以通过教务处官网提交申请表。",

"登录教务系统即可查看个人成绩。",

"校团委网站有详细的社团介绍。",

"携带身份证到学工部办理。"

]

}

# 将问题和答案保存为列表

questions = knowledge_base["questions"]

answers = knowledge_base["answers"]

# 使用TF-IDF向量化

vectorizer = TfidfVectorizer()

tfidf_matrix = vectorizer.fit_transform(questions)

# 用户输入

user_input = input("请输入你的问题：")

# 向量化用户输入

user_vector = vectorizer.transform([user_input])

# 计算相似度

similarities = cosine_similarity(user_vector, tfidf_matrix).flatten()

# 找出最相似的问题

most_similar_index = similarities.argmax()

# 输出答案

print("答案：", answers[most_similar_index])

import pymysql

from sklearn.feature_extraction.text import TfidfVectorizer

from sklearn.metrics.pairwise import cosine_similarity

# 连接数据库

conn = pymysql.connect(

host='localhost',

user='root',

password='your_password',

database='campus_qa'

)

cursor = conn.cursor()

# 查询知识库数据

cursor.execute("SELECT question, answer FROM qa_table")

results = cursor.fetchall()

# 提取问题和答案

questions = [row[0] for row in results]

answers = [row[1] for row in results]

# 向量化

vectorizer = TfidfVectorizer()

tfidf_matrix = vectorizer.fit_transform(questions)

# 用户输入

user_input = input("请输入你的问题：")

# 向量化

user_vector = vectorizer.transform([user_input])

# 相似度计算

similarities = cosine_similarity(user_vector, tfidf_matrix).flatten()

# 找出最相似的问题

most_similar_index = similarities.argmax()

# 输出答案

print("答案：", answers[most_similar_index])

# 关闭连接

cursor.close()

conn.close()

from transformers import BertTokenizer, BertForSequenceClassification

import torch

# 加载预训练的BERT模型和分词器

model_name = 'bert-base-uncased'

tokenizer = BertTokenizer.from_pretrained(model_name)

model = BertForSequenceClassification.from_pretrained(model_name)

# 定义知识库

knowledge_base = {

"questions": [

"如何申请奖学金？",

"怎么查询成绩？",

"学校有哪些社团？",

"如何办理学生证？"

],

"answers": [

"您可以通过教务处官网提交申请表。",

"登录教务系统即可查看个人成绩。",

"校团委网站有详细的社团介绍。",

"携带身份证到学工部办理。"

]

}

# 用户输入

user_input = input("请输入你的问题：")

# 分词并编码

inputs = tokenizer(user_input, return_tensors="pt", padding=True, truncation=True)

# 获取嵌入向量

with torch.no_grad():

outputs = model(**inputs)

embeddings = outputs.last_hidden_state.mean(dim=1)

# 对知识库中的问题进行编码

question_embeddings = []

for question in knowledge_base["questions"]:

q_inputs = tokenizer(question, return_tensors="pt", padding=True, truncation=True)

with torch.no_grad():

q_outputs = model(**q_inputs)

q_embeddings = q_outputs.last_hidden_state.mean(dim=1)

question_embeddings.append(q_embeddings)

# 计算相似度

similarities = [torch.nn.functional.cosine_similarity(embeddings, q_emb) for q_emb in question_embeddings]

most_similar_index = similarities.index(max(similarities))

# 输出答案

print("答案：", knowledge_base["answers"][most_similar_index])