Research And Construction Of Causal Knowledge Base

The rapid development of Internet technology, and constantly promote the change of the way of thinking of mankind; the explosive growth of data, and constantly changing the pattern of the binary world.Today, the rapid expansion of data information, has brought us into the era of big data。Data information has become the main points of the lifeline to grasp the key points of the industry. Data driven information technology revolution is subverting the human life as in an unprecedented fashion.Data and information are interrelated, the data after mining and processing can become the information that people need. In the era of big data, there is great value in the data, extracting useful information from the massive data, identifying the patterns, laws, and the correlation between them. It is a new way for human to understand the world. In social media, people’s daily behavior and emotional statement contains a lot of life experience summary, including a large number of common sense causal relationship. Under the background of big data, according to the existing experience knowledge from the text data to identify the causal relationship between things, summed up the causal relationship between the accident, to build a causal relationship between the knowledge base(causal network). The prediction function of computer system is replaced by the knowledge of common sense thinking in people’s life, and even the reasoning that the human mind can not reach the deep cause and effect relationship, To make it clear the interrelation and adjust the negative factors.This paper puts forward a method of extracting relations within the sentence causal, based on knowledge of Chinese language and literature on the use of causality cue word recognition with show causality of causal sentences; according to the grammatical structure and syntactic dependency relation induces syntactic pattern matching rules, in recognition of the causal relationship between entities; at the same time according to the degree adverb the sentence contains emotion words and negative words and syntactic patterns to determine causal strength between entities; according to semantic similarity, causal relationship between the entity of semantic similarity fusion, the causal relationship between entity sets to form; and calculate the causal entity to the co-occurrence frequency of causal support.Building the knowledge base of causal relationship.According to the extraction of the causal relationship in the text, the paper puts forward the following innovative points:(1) in the extraction of emotional words, words, negative words, the reasons for the analysis of the extent to which part of the results of some of the impact, that is, the calculation of the intensity of causality;(2) calculate the causal relationship support by calculating the co-occurrence frequency of causal entities in the data source;(3) cluster operations on the basis of semantic similarity by using the semantic similarity, and transform the causal data into causal knowledge.