科技通报
科技通報
과기통보
Bulletin of Science and Technology
2015年
10期
187-189
,共3页
B2C模式%忠诚度%特征挖掘
B2C模式%忠誠度%特徵挖掘
B2C모식%충성도%특정알굴
B2C model%loyalty%characteristics mining
在B2C模式下对客户的忠诚度特征挖掘是对商户交易信任度评价的基础,进而促进用户忠诚度的提升.传统的B2C忠诚度特征挖掘算法采用基于负反馈动态的渐进控制算法,当用户数据呈现稀疏状特征时,挖掘效果不好.提出一种基于多模判决反馈的特征挖掘算法以提升B2C忠诚度.构建B2C模式下的忠诚度特征分析评估机制,根据B2C模式下多模判决反馈忠诚度控制稳定性理论,提出一种电子商务用户忠诚度评价渐进提升控制算法,这里选取Lyapunov函数构建提升目标函数,实现特征挖掘算法改进.仿真结果表明,采用该算法进行B2C模式下的商户忠诚度特征挖掘和评价,忠诚度评价准确度有所提高,用户的数据信息特征预测性能较好,特征挖掘精度提升,提高了对B2C模式下的商户的定量评价精度.
在B2C模式下對客戶的忠誠度特徵挖掘是對商戶交易信任度評價的基礎,進而促進用戶忠誠度的提升.傳統的B2C忠誠度特徵挖掘算法採用基于負反饋動態的漸進控製算法,噹用戶數據呈現稀疏狀特徵時,挖掘效果不好.提齣一種基于多模判決反饋的特徵挖掘算法以提升B2C忠誠度.構建B2C模式下的忠誠度特徵分析評估機製,根據B2C模式下多模判決反饋忠誠度控製穩定性理論,提齣一種電子商務用戶忠誠度評價漸進提升控製算法,這裏選取Lyapunov函數構建提升目標函數,實現特徵挖掘算法改進.倣真結果錶明,採用該算法進行B2C模式下的商戶忠誠度特徵挖掘和評價,忠誠度評價準確度有所提高,用戶的數據信息特徵預測性能較好,特徵挖掘精度提升,提高瞭對B2C模式下的商戶的定量評價精度.
재B2C모식하대객호적충성도특정알굴시대상호교역신임도평개적기출,진이촉진용호충성도적제승.전통적B2C충성도특정알굴산법채용기우부반궤동태적점진공제산법,당용호수거정현희소상특정시,알굴효과불호.제출일충기우다모판결반궤적특정알굴산법이제승B2C충성도.구건B2C모식하적충성도특정분석평고궤제,근거B2C모식하다모판결반궤충성도공제은정성이론,제출일충전자상무용호충성도평개점진제승공제산법,저리선취Lyapunov함수구건제승목표함수,실현특정알굴산법개진.방진결과표명,채용해산법진행B2C모식하적상호충성도특정알굴화평개,충성도평개준학도유소제고,용호적수거신식특정예측성능교호,특정알굴정도제승,제고료대B2C모식하적상호적정량평개정도.
In the B2C model of customer loyalty characteristics of mining is the basis of merchant transaction trust degree evaluation, and promote the customer loyalty to enhance. Characteristics of B2C loyalty traditional mining algorithms using negative feedback control algorithm based on evolutionary dynamics, when the user data show a sparse shape characteris-tic, mining effect is not good. A mining algorithm of feature B2C loyalty multimode decision feedback based on lifting is pro-posed. Analysis of the construction of evaluation mechanism of loyalty characteristics under the mode of B 2C, according to B2C mode multimode decision feedback loyalty control stability theory, put forward a kind of evaluation of e-commerce us-er loyalty gradual lifting control algorithm, here's a selection of Lyapunov function to construct target to improve function, realize the characteristics of the improved algorithm. The simulation results show that, the merchant loyalty feature mining and evaluation using the algorithm of B2C mode, loyalty evaluation accuracy is improved, the data information characteris-tics of users better prediction performance, feature mining precision improvement, improve the quantification of the B 2C mode of the merchants of evaluation accuracy.