Research On Intelligence Test Paper Algorithm Based On Swarm Intelligence Optimization

Nowdays,most of intelligent test paper systems have defects of low success rate or low speed of test paper.In the intelligent test paper systems,the important part is the design of the intelligent test paper algorithm when considering how to generate papers meeting the needs.Therefore,it is helpful that the analysis,the research and the improvement of the intelligent test paper algorithm to promote the teaching quality.The traditional intelligent test paper algorithms are usually on the basis of the random algorithm and item response theory,but these algorithms have weaknesses of low success rate and the consumption of manpower and material resources.Therefore,on the basic theory of intelligent test paper algorithm,the bacteria foraging optimization algorithm based on the cloud model(CBFOA)and the adaptive fruit fly optimization algorithm based on normal cloud model(CAFOA)are presented and applied to generate papers integrating with the thought of bacteria foraging optimization algorithm and fruit fly optimization algorithm.The results of experiment show that these test paper algorithms are good for generating papers.The major works are listed below:(1)The representative intelligent test paper algorithms are analyzed,the advantages and disadvantages are summarized,and the similarities and differences of intelligent test paper algorithms are compared between genetic algorithm and swarm intelligent algorithms.At last,the significance is summarized on BFOA and FOA.(2)The bacteria foraging optimization algorithm based on the cloud model is presented.Based on the standard of bacteria foraging optimization algorithm and cloud model,firstly,in the operation of chemotaxis and reproduction,the conception of sensitivity is given and adjusted by the X-conditional cloud generator for controlling swim steps,combined with the characters of randomness and stability of the cloud model.Then,in the operation of elimination and dispersal,the adaptive and non-linear probability of elimination and dispersal is adopted by the forward normal cloud generator.Finally,this algorithm is used to generate papers,compared and analyzed with the experiment of Genetic Algorithm.The results of experiment show that this algorithm is better than Genetic Algorithm both in convergence rate and quality of optimization,which provides possibility between the reduction of time and the improvement of success rate of test paper.(3)The adaptive fruit fly optimization algorithm based on normal cloud model is presented.On the basis of fruit fly optimization algorithm,firstly,the conception of sensitive factor is introduced and adjusted adaptively for controlling search step to update location.Then,the randomness and fuzziness of smell concentration parameter is described by normal cloud model and adjusted to finish osphresis search operation automatically,and the detailed steps of CAFOA are given.Finally,this algorithm is used to the automatic test,compared and analyzed with the experiment of FOA.The results of experiment show that this algorithm has better advantages of test efficiency and accuracy.The research enriches and completes the intelligence test paper algorithm based on swarm intelligence optimization,and provides the theoretical basis for the intelligence test paper.