基于LDA-GA算法的移动目录优化研究

Research on Mobile-oriented Catalog Optimization Based on LDA-GA Algorithm

针对移动设备向用户推荐产品时受限于尺寸的问题,目前普遍采用个性化协作推荐算法来实现开发面向移动目录(MOC),但是传统的方法存在大数据环境下适应度不高、协作能力差等不足。为解决此问题,首先将主题建模算法与遗传算法相结合开发出LDA-GA算法,然后设计富有吸引力和协作性的产品推荐目录,最后将MOC应用在亚马逊APP和淘宝网APP进行实验比对分析并进行优化。实验结果表明:LDA-GA算法面对大量用户和产品数据时移动目录适应度更高、协作性更强,客户受众面大,推介效果更好。

To solve the problem that mobile devices are subject to the limitation of sizes when recommending products to users,the individualized collaboration is recommended to realize the development of mobile-ori-ented catalog （MOC）. However, traditional methods are of low adaptability and poor collaboration under the environment of big data. To solve the problem,the topic modeling algorithm and genetic algorithm are firstly combined to develop a LDA-GA algorithm; secondly, the attractive and collaborative recommenda-tion catalog is designed; finally,MOC is applied in Amazon APP and Taobao. com APP for comparative a-nalysis and optimization. The experimental results show that the LDA-GA algorithm is more adaptive, more cooperative and more effective in the environments of a large number of users and product data.