TAO Xiaxin - Associate research professor, Institute of Engineering Mechanics, Harbin, CHINA
DU Wei - Graduate student, Institute of Engineering Mechanics, Harbin, CHINA
ZHENG Guangfen - Research associate,Institute of Engineering Mechanics, Harbin, CHINA

A software for estimating the seismicity parameters with considering the tendency in a seismic province or zone is presented in this paper. The seismicity rhythm is recognized by means of the Artificial Neural Network approach, which becomes popular in Artificial In-telligence field. The estimation of seismicity tendency in future years is carried out in two procedures: ۱.The seismic history is characterized by four stages: energy accumulation, increasing release, main release and remnant release stage; ۲.The seismicity is character-ized directly by the frequencies of earthquakes with various magnitudes. After learning from the earthquake catalog of the southeast coast zone of China, the net-work memorized the rhythm successfully. In order to evaluate the reliability of the fore-casting by the software, the historical data in recent ۱۰۰ years, and then ۱۷۹۰-۱۹۹۰ and ۱۶۹۰-۱۹۹۰ were withdrawn alternatively, the network was trained again, the results fit the actual situation quite well.