手机资费套餐推荐方法及系统的初步实现

发布时间：2018-01-05 08:00

本文关键词：手机资费套餐推荐方法及系统的初步实现　出处：《东北大学》2013年硕士论文　论文类型：学位论文

【摘要】：随着电信市场重组,以及开展全业务运营和3G牌照的发放,中国通信市场发生了深刻变革,市场竞争日益激烈,各运营商都十分迫切地希望从各个方面提升对客户的精细化服务水平,提高客户满意度。手机资费套餐作为通信行业进行客户获取、价值提升、维系挽留的一个重要工具。如何充分利用这一工具,既让客户满意,又使企业从中获益,是当前亟需解决的问题。因此,从用户需求的角度出发,找寻科学的手机资费套餐推荐方式,设计合理高效的手机资费套餐推荐系统显得尤为重要。基于此背景,本文主要研究了如何在用户消费行为多样化的前提下,为用户提供更为人性化的准确的手机资费套餐推荐的问题。从用户的角度出发,并参考一般的推荐方法,主要提出了三类针对于不同消费习惯人群的手机资费套餐推荐方法。首先,针对由于职业性质而影响手机消费情况的用户群,通过实例详细阐述了以数据分类技术作为理论基础的数据分类处理过程,及各个手机资费套餐在各分类中的权重取值规则,并且通过寻找最小邻居来解决无法确定用户数据类型时的分类过程,设计了以职业种类与收入水平为推荐条件的推荐界面,提出了根据用户类型而推荐该类型下权重较高的手机资费套餐的推荐方式。然后,针对存在手机资费套餐历史消费数据并希望了解自身消费习惯的用户群,以二次平滑预测方法作为理论依据,设计了通过用户输入手机号而获得用户历史消费数据的推荐界面,提出了根据预测结果为用户进行手机资费套餐推荐的推荐方式。接着,针对无法明确给出每月准确消费量的用户群,提出了如下两种模糊推荐方法。(1)建立了基于区间模糊数理论的数据处理方法,设计了针对手机资费套餐各个属性的最小消费量、最大消费量及对消费属性语义评价为推荐条件的用户推荐界面,并且通过实例进行了推荐说明。(2)以模糊四则运算为基础建立了现有手机资费套餐相对于用户需求的基本消费的具体算法,设计了以三角模糊形式进行用户需求输入的推荐界面,分析了可能出现的手机资费套餐模糊费用的隶属函数种类,并且以九种模糊排序方法为基础对手机资费套餐推荐实例进行了模糊排序比较。最后,在所提出的手机资费套餐推荐方法的基础上,应用Flash Builder和Visual Studio开发平台,结合SQL Server数据库,设计并初步实现了手机资费套餐推荐系统。该系统包含推荐模块、排行榜及资费详情等主要功能模块。
[Abstract]:With the reorganization of the telecommunications market, as well as the development of full-service operations and 3G license issuance, China's telecommunications market has undergone profound changes, the market competition is becoming increasingly fierce. All operators are eager to improve the level of service to customers from all aspects to improve customer satisfaction. Mobile phone tariff package as the communications industry to obtain customers, the value of promotion. How to make full use of this tool not only to make customers satisfied, but also to make enterprises benefit from it, is an urgent problem. Therefore, from the point of view of user needs, how to make full use of this tool. It is very important to find a scientific way to recommend mobile phone tariff package, and to design a reasonable and efficient mobile phone tariff package recommendation system. Based on this background, this paper mainly studies how to diversify the consumer behavior. To provide users with a more humane and accurate mobile phone tariff package recommendation. From the user's point of view, and refer to the general recommended methods. Mainly put forward three kinds of mobile phone tariff packages recommended methods for different consumption habits. First of all, for the professional nature of mobile phone consumption affecting the user group. Through an example, the data classification process based on data classification technology and the weight rules of each mobile phone tariff package in each classification are expounded in detail. And by looking for the smallest neighbor to solve the classification process when the user data type can not be determined, a recommendation interface based on occupational type and income level is designed. According to the user type and recommend the high weight of the mobile phone tariff package. Then, aiming at the existence of historical consumption data and want to know their own consumption habits of users. Based on the quadratic smoothing prediction method, the user input mobile phone number to obtain the user history consumption data recommendation interface is designed. According to the forecast results for users to recommend the mobile phone tariff package. Then, for the users who can not clearly give the monthly accurate consumption. Put forward the following two fuzzy recommendation methods. 1) establish the data processing method based on interval fuzzy number theory, and design the minimum consumption for each attribute of mobile phone tariff package. The user recommendation interface with the maximum consumption and semantic evaluation of the consumption attribute as the recommendation condition. And through an example to recommend the description. 2) based on the fuzzy four principles operation to establish the existing mobile phone tariff set relative to the user needs of the basic consumption of the specific algorithm. The recommended interface for user requirement input in the form of triangle fuzzy is designed, and the possible membership function types of mobile phone tariff set fuzzy cost are analyzed. And on the basis of nine kinds of fuzzy sorting methods, the paper makes a comparison of the mobile phone tariff package recommendation examples. Finally, on the basis of the proposed mobile phone tariff package recommendation method. The application of Flash Builder and Visual Studio development platform, combined with SQL Server database. A mobile phone tariff package recommendation system is designed and implemented, which includes the recommendation module, ranking list and tariff details and other main functional modules.
【学位授予单位】：东北大学
【学位级别】：硕士
【学位授予年份】：2013
【分类号】：F626

【相似文献】