KGQA

基于知识图谱的金融数据问答系统

Knowledge Graph Question Answering with Chinese Financial data from TuShare

运行实例：

• Stored financial data as knowledge graph with neo4j
• Extract question type & entity from user query with Aho–Corasick algorithm
• Detect if query is random chat using fasttext

Step

run semantic_parser.py to build a AC automation trie tree

run classifier.py to build a fasttext classifier

run store_to_neo4j.py transform the data into neo4j

run main.py

Overall workflow:

neo4j

You can download neo4j at https://neo4j.com/download/

create a new DBMS in local, and then Start the server, so you can dump data into KG and visualize the data

AC自动机: Aho-Corasick

Input: text = "ahishers"    
       arr[] = {"he", "she", "hers", "his"}

Output:
   Word his appears from 1 to 3
   Word he appears from 4 to 5
   Word she appears from 3 to 5
   Word hers appears from 4 to 7

RUNTIME: O(n + m + z)

AC自动机的基础是Trie树。和Trie树不同的是，树中的每个结点除了有指向孩子的指针（或者说引用），还有一个fail指针，它表示输入的字符与当前结点的所有孩子结点都不匹配时(注意，不是和该结点本身不匹配)，自动机的状态应转移到的状态（或者说应该转移到的结点）。