Objective: To study the screening methods of core herbal combinations of anti-Alzheimer prescriptions by using Traditional Chinese Medicine Inheritance System (TCMIS). Method: The traditional medical prescriptions for anti-Alzheimer in Dictionary of Chinese Medicine Prescriptions (DCMP) were selected and TCMIS V2.5 was used to input the prescription information and construct the database. The commonly used herbs

combination principles

Dominance & Recessive herb pairs and the correlations between prescriptions and symptoms were analyzed and new prescriptions were developed with data mining methods such as association rules

improved mutual information method and complex system entropy clustering. Result: The 30 commonly used herbs were concluded based on the statistic frequency analysis of 102 anti-Alzheimer prescriptions. Ginseng Radix et Rhizoma

Polygalae Radix

Acori Poria and Tatarinowii Rhizoma ranked top four in frequency. The 20 commonly used herb pairs were concluded based on the association rules analysis of 102 anti-Alzheimer prescriptions. Ginseng Radix et Rhizoma-Polygalae Radix

Polygalae Radix-Tatarinowii Rhizoma

and Ginseng Radix et Rhizoma and Tatarinowii Rhizoma ranked top three among the commonly used herb pairs. In accordance with the correlations between prescriptions and symptoms

the core herbal combinations for anti-Alzheimer prescriptions in DCMP were greatly correlated with Ginseng Radix et Rhizoma

Polygalae Radix

Tatarinowii Rhizoma and Acori Poria. The 40 new core combinations and 2 newly developed prescriptions were concluded based on improved mutual information method and complex system entropy clustering in 102 prescriptions for anti-Alzheimer. Conclusion: The ingredients in core combinations of anti-Alzheimer prescriptions were highly coincided with those in Kaixin San in DCMP

with clear treatment significance for memory loss. The two newly developed prescriptions for anti-Alzheimer found by TCMIS V2.5 have some clinical value.