WO2020147265A1 - 一种基于多源信息融合的移动电子商务推荐方法和系统 - Google Patents
一种基于多源信息融合的移动电子商务推荐方法和系统 Download PDFInfo
- Publication number
- WO2020147265A1 WO2020147265A1 PCT/CN2019/092977 CN2019092977W WO2020147265A1 WO 2020147265 A1 WO2020147265 A1 WO 2020147265A1 CN 2019092977 W CN2019092977 W CN 2019092977W WO 2020147265 A1 WO2020147265 A1 WO 2020147265A1
- Authority
- WO
- WIPO (PCT)
- Prior art keywords
- evidence
- original
- data
- information
- evidence body
- Prior art date
Links
- 238000000034 method Methods 0.000 title claims abstract description 53
- 230000004927 fusion Effects 0.000 title claims abstract description 39
- 238000013528 artificial neural network Methods 0.000 claims abstract description 20
- 230000008569 process Effects 0.000 claims abstract description 10
- 238000001228 spectrum Methods 0.000 claims abstract description 10
- 238000007781 pre-processing Methods 0.000 claims abstract description 7
- 230000015572 biosynthetic process Effects 0.000 claims description 8
- 238000003786 synthesis reaction Methods 0.000 claims description 8
- 210000002569 neuron Anatomy 0.000 claims description 6
- 238000012546 transfer Methods 0.000 claims description 6
- 239000008186 active pharmaceutical agent Substances 0.000 claims description 5
- 230000004044 response Effects 0.000 claims description 5
- 238000005311 autocorrelation function Methods 0.000 claims description 3
- 230000003595 spectral effect Effects 0.000 claims description 3
- 238000000053 physical method Methods 0.000 claims 1
- 230000006399 behavior Effects 0.000 description 4
- 238000012545 processing Methods 0.000 description 4
- 238000011160 research Methods 0.000 description 4
- 230000009286 beneficial effect Effects 0.000 description 2
- 238000007418 data mining Methods 0.000 description 2
- 230000007547 defect Effects 0.000 description 2
- 238000001914 filtration Methods 0.000 description 2
- 238000005065 mining Methods 0.000 description 2
- 230000008859 change Effects 0.000 description 1
- 230000008094 contradictory effect Effects 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000007123 defense Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 230000010354 integration Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000003909 pattern recognition Methods 0.000 description 1
- 238000012552 review Methods 0.000 description 1
- 230000002194 synthesizing effect Effects 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N3/00—Computing arrangements based on biological models
- G06N3/02—Neural networks
- G06N3/04—Architecture, e.g. interconnection topology
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/06—Buying, selling or leasing transactions
Definitions
- the invention relates to the technical field of information fusion, in particular to a mobile e-commerce recommendation method and system based on multi-source information fusion.
- recommendation system In order to solve the above problems, a recommendation system was invented and widely used by e-commerce practitioners. It has now become an important research topic in information science and decision support systems. At present, the research of recommendation systems generally includes content-based filtering (CBF), collaborative filtering (CF) and other data mining techniques, such as decision trees, association rules, and semantic methods.
- CBF content-based filtering
- CF collaborative filtering
- other data mining techniques such as decision trees, association rules, and semantic methods.
- the e-commerce recommendation system is not deep enough to mine consumers' online behaviors in multi-source mining.
- the recommendation system only pays attention to information about products and consumers' shopping behaviors on the shopping platform. Therefore, the accuracy of recommendations is limited.
- the existing mobile e-commerce recommendation system does not integrate user location information, and the coverage of recommendations is limited.
- the purpose of the present invention is to provide a mobile e-commerce recommendation method and system based on multi-source information fusion.
- the present invention can effectively increase recommendation accuracy, indirectness and coverage rate, and improve recommendation service performance.
- the present invention proposes a mobile e-commerce recommendation method based on multi-source information fusion, the method includes:
- S1 Obtain user information and corresponding consumption data from multiple information sources, and preprocess the obtained user information and corresponding consumption data to obtain original evidence bodies, which are divided into several categories;
- S2 Calculate the evidence weight of each type of data in the original evidence body based on the original evidence body through a radial basis function and a neural network algorithm, and the evidence weight is used to distinguish the recommended value of each type of data;
- step S1 obtaining user information and corresponding consumption data from multiple information sources includes:
- nickname-mobile phone number to extract user characteristic data and user's high-frequency attention data
- platform ID-mobile phone number to mine users' high-frequency search data and historical purchase information from the databases of various shopping platforms.
- preprocessing the acquired user information and corresponding consumption data includes:
- step S2 based on the original evidence body, the method of calculating the evidence weight of each type of data in the original evidence body through a radial basis function and a neural network algorithm includes:
- the radial basis function neural network is used to calculate the evidence weight of each type of data in the original evidence body
- the hidden layer of the radial basis function neural network is a radial basis Gaussian layer, and the corresponding transfer function is a Gaussian function:
- the output layer of the radial basis function neural network is the softmax layer, and the corresponding transfer function is:
- c is the number of output categories
- x is the neuron input
- y is the neuron output
- ⁇ j is the variance
- j is the number of inputs.
- step S3 according to the evidence weight, the method of using D-S evidence theory to perform information fusion on the original evidence body to obtain a new evidence body includes:
- the feasibility of evidence ⁇ is introduced as a criterion to modify the original evidence body M, and the basic probability distribution value of the modified original evidence body M is calculated according to the following formula:
- m(A m ) ⁇ m′(A m )
- m 1,2,...,n ⁇ , m′(A m ) ⁇ M′;
- the modified original evidence body M is fused to obtain a new evidence body, and the following formula is used to calculate the basic probability distribution value of the new evidence body after fusion:
- A, B, C, D represent evidence
- m represents the synthesis rule of evidence
- j represents the number of evidence in the evidence body
- ⁇ represents the identification frame
- k represents the degree of conflict between the evidence
- the influencing factor of the evidence feasibility ⁇ includes one or more of the timeliness, stability and comprehensiveness of the information data.
- the method further includes:
- the new evidence is normalized.
- step S4 using a power spectrum estimation method to process the new evidence body to obtain a recommendation decision means:
- the present invention also mentions a mobile e-commerce recommendation system based on multi-source information fusion, which includes:
- the module is used to process the new evidence body by adopting the power spectrum estimation method to obtain the recommended decision-making module.
- Fig. 1 is a flowchart of the mobile e-commerce recommendation method based on multi-source information fusion of the present invention.
- the present invention proposes a mobile e-commerce recommendation method based on multi-source information fusion, and the method includes:
- S1 Obtain user information and corresponding consumption data from multiple information sources, and preprocess the obtained user information and corresponding consumption data to obtain original evidence bodies, which are divided into several categories.
- Information sources include cloud data, databases of various shopping platforms, web page information and video information browsed by users, user data provided by search engines, user chat record data, location information, voices, pictures, videos, etc. stored on user mobile terminals .
- the present invention proposes that, preferably, information is obtained from all information sources that may obtain user information to fully understand the needs of consumers, and then recommend products according to the needs. Especially for location information, there is a close relationship between the location of consumers and their shopping behavior.
- the mobile e-commerce recommendation system can choose to integrate user location information, that is, a method of integrating multiple sources of information Introduced into the integration of location and historical behavior information.
- Obtaining methods include the following: through "nickname-mobile phone number” to extract user characteristic data and user's high-frequency attention data, through “platform ID-mobile phone number” to mine users' high-frequency search data from the databases of various shopping platforms And historical purchase information and more.
- the invention aims at multi-source information fusion and decision-making recommendation in mobile e-commerce, using location information, social platform comments, and product information and user information to effectively increase recommendation accuracy.
- This method makes up for the accuracy defect that the previous recommendation method only focuses on the product and the user's shopping information on the consumer platform without combining the user's other information.
- preprocessing the acquired user information and corresponding consumption data includes:
- the acquired data format is unified, which is convenient for users to read and/or machine processing.
- S2 Calculate the evidence weight of each type of data in the original evidence body based on the original evidence body through a radial basis function and a neural network algorithm, and the evidence weight is used to distinguish the recommended value of each type of data.
- the present invention uses an improved radial basis function neural network to calculate the evidence weight of each type of data in the original evidence body.
- the hidden layer of the radial basis function neural network is a radial basis Gaussian layer, and the corresponding transfer function is a Gaussian function:
- the output layer of the radial basis function neural network is the softmax layer, and the corresponding transfer function is:
- c is the number of output categories
- x is the neuron input
- y is the neuron output
- ⁇ j is the variance
- j is the number of inputs.
- m(A m ) ⁇ m′(A m )
- m 1,2,...,n ⁇ , m'(A m ) ⁇ M'.
- the factors affecting the feasibility of the evidence ⁇ include one or more of the timeliness, stability and comprehensiveness of the information data.
- the modified original evidence body M is fused according to the synthesis rules based on the local conflict allocation strategy to obtain a new evidence body.
- the following formula is used to calculate the fusion of the new evidence body.
- A, B, C, D represent evidence
- m represents the synthesis rule of evidence
- j represents the number of evidence in the evidence body
- ⁇ represents the identification frame
- k represents the degree of conflict between the evidence
- the combination of using feasibility to modify the evidence body and assigning and combining rules based on local conflicts can not only improve the accuracy of recommendations, reduce the response time of the recommendation system, but also increase the coverage of recommendations.
- the new evidence is normalized to facilitate subsequent processing.
- the present invention also mentions a mobile e-commerce recommendation system based on multi-source information fusion.
- the system includes the following modules:
Landscapes
- Engineering & Computer Science (AREA)
- Physics & Mathematics (AREA)
- Business, Economics & Management (AREA)
- Theoretical Computer Science (AREA)
- Accounting & Taxation (AREA)
- Finance (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Data Mining & Analysis (AREA)
- Strategic Management (AREA)
- Marketing (AREA)
- Economics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Development Economics (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Business, Economics & Management (AREA)
- Evolutionary Computation (AREA)
- General Health & Medical Sciences (AREA)
- Molecular Biology (AREA)
- Computing Systems (AREA)
- General Engineering & Computer Science (AREA)
- Mathematical Physics (AREA)
- Software Systems (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
本发明公开了一种基于多源信息融合的移动电子商务推荐方法,包括:S1:从多个信息源获取用户信息和对应的消费数据,对获取的用户信息和对应的消费数据进行预处理,以获取原始证据体,所述原始证据体被划分成若干类;S2:基于所述原始证据体,通过径向基函数和神经网络算法,计算原始证据体中每类数据的证据权重,所述证据权重用于区分每类数据的推荐价值;S3:根据所述证据权重,运用D-S证据理论对原始证据体进行信息融合,以获得新的证据体;S4:采用功率谱估计方法对所述新的证据体进行处理,以获取推荐决策。本发明能够有效增加推荐准确性、间接性和覆盖率,提高推荐服务性能。
Description
本发明涉及信息融合技术领域,具体而言涉及一种基于多源信息融合的移动电子商务推荐方法和系统。
随着用户消费习惯的变化,移动电子商务已成为一种趋势。然而,连续生成大量数据不仅对于消费者在搜索有意义的产品时不方便,而且还意味着很少购买一些产品。缺乏对用户和资源信息的深入挖掘已成为制约移动商务推荐系统预测分析的主要瓶颈。
为了解决上述问题发明了推荐系统,并被电子商务从业者广泛应用,现已成为信息科学和决策支持系统的重要研究课题。目前,推荐系统的研究一般包括基于内容的过滤(CBF),协同过滤(CF)和其他数据挖掘技术,如决策树,关联规则和语义方法。
现有的电子商务推荐系统存在着两大问题。首先,电子商务推荐系统不够深入,无法在多源挖掘中挖掘消费者的在线行为。推荐系统只关注产品和消费者在购物平台上的购物行为的信息,因此,推荐的准确性是有限的。其次,现有的移动电子商务推荐系统没有整合用户位置信息,推荐的覆盖性是有限的。
Yager、Gregor和许多其他学者深入研究了多源信息融合框架,信息分类,自动推理,异构数据处理,云计算和点对点(P2P)信息融合的网络信任模型。大多数信息融合模型都是基于美国国防部建立的JDL模型,它实现了从四个不同处理层面融合多源信息的要求。随着多源信息融合技术研究的发展,它已被用于模式识别,数据挖掘,知识发现等方面。然而,基于移动电子商务推荐系统中的位置的多源信息融合的研究较少。
发明内容
本发明目的在于提供一种基于多源信息融合的移动电子商务推荐方法和系统,首先根据消费平台内外的两种类型信息,获取用户信息和数据;其次通过径向基函数和神经网络算法,计算推荐证据权重;再运用D-S证据理论,采用可行度来修改证据体,进行信息融合;最后综合考虑信息的时效性,稳定性,全面性等众多因素,采用功率谱估计方法处理融合结果,最终得到推荐决策。本发明能够有效增加推荐准确性、间接性和覆盖率,提高推荐服务性能。
为达成上述目的,结合图1,本发明提出一种基于多源信息融合的移动电子商务推荐方法,所述方法包括:
S1:从多个信息源获取用户信息和对应的消费数据,对获取的用户信息和对应的消费数据进行预处理,以获取原始证据体,所述原始证据体被划分成若干类;
S2:基于所述原始证据体,通过径向基函数和神经网络算法,计算原始证据体中每类数据的证据权重,所述证据权重用于区分每类数据的推荐价值;
S3:根据所述证据权重,运用D-S证据理论对原始证据体进行信息融合,以获得新的证据体;
S4:采用功率谱估计方法对所述新的证据体进行处理,以获取推荐决策。
进一步的实施例中,步骤S1中,从多个信息源获取用户信息和对应的消费数据包括:
通过“昵称-手机号码”以提取用户特征数据以及用户的高频关注数据,通过“平台ID-手机号码”以从各个购物平台的数据库中挖掘用户的高频搜索数据与历史购买信息。
进一步的实施例中,步骤S1中,对获取的用户信息和对应的消费数据进行预处理包括:
采用微格式对获取的用户信息和对应的消费数据进行表示、存储、集成和管理。
进一步的实施例中,步骤S2中,基于所述原始证据体,通过径向基函数和神经网络算法,计算原始证据体中每类数据的证据权重的方法包括:
采用径向基神经网络以计算原始证据体中每类数据的证据权重;
其中,c为输出分类数目,x为神经元输入,y为神经元输出,σ
j为方差,j为输入个数。
进一步的实施例中,步骤S3中,根据所述证据权重,运用D-S证据理论对原始证据体进行信息融合,以获得新的证据体的方法包括:
根据D-S证据理论,引入证据可行度μ作为评判标准以修改原始证据体M,根据下述公式计算修改后的原始证据体M的基本概率分配值:
M=[m(A
1)m(A
2)…m(A
n)m(Θ)]
根据基于局部冲突分配策略的合成规则,对修改后的原始证据体M进行融合,以获得新的证据体,采用下述公式计算融合后所得新证据体的基本概率分配值:
进一步的实施例中,所述证据可行度μ的影响因子包括信息数据的时效性、稳定性和全面性中的一种或者多种。
进一步的实施例中,所述方法还包括:
响应于每两个原始证据合成新的证据,对新的证据进行归一化处理。
进一步的实施例中,步骤S4中,采用功率谱估计方法对所述新的证据体进行处理,以获取推荐决策是指,
根据用户需求和特征信息x(n)的自相关函数
代表任 意两个不同用户在不同时刻的用户需求和特征信息x
N(n)与x
N(n+m)之间的相关程度,由该相关程度计算出功率谱密度P
xx(ω)=∑
mΦ
xx(m)e
-zωm,与相关程度为一对傅里叶变换。
基于前述方法,本发明还提及一种基于多源信息融合的移动电子商务推荐系统,所述系统包括:
用以从多个信息源获取用户信息和对应的消费数据,对获取的用户信息和对应的消费数据进行预处理,以获取原始证据体的模块,所述原始证据体被划分成若干类;
用以基于所述原始证据体,通过径向基函数和神经网络算法,计算原始证据体中每类数据的证据权重的模块,所述证据权重用于区分每类数据的推荐价值;
用以根据所述证据权重,运用D-S证据理论对原始证据体进行信息融合,以获得新的证据体的模块;
用以采用功率谱估计方法对所述新的证据体进行处理,以获取推荐决策的模块。
以上本发明的技术方案,与现有相比,其显著的有益效果在于:
1)根据消费平台内外的两种类型信息,获取用户信息和数据,确保数据来源多样化。
2)结合径向基神经网络与D-S证据理论,使用位置信息,社交平台评论,产品信息和用户信息,有效增加推荐准确性,弥补了以往推荐方法仅关注产品以及用户在消费平台的购物信息,不结合用户的其他信息的准确性缺陷;
3)采用可行度来修改证据体,进行信息融合,避免D-S理论的鲁棒性和一票否决性;
4)采用可行度来修改证据体、以及根据局部冲突分配合成规则这两个方面的结合,既能提高推荐的准确性,降低推荐系统响应时间,又能增加推荐覆盖率,提高推荐服务性能。
应当理解,前述构思以及在下面更加详细地描述的额外构思的所有组合只要在这样的构思不相互矛盾的情况下都可以被视为本公开的发明主题的一部分。另外,所要求保护的主题的所有组合都被视为本公开的发明主题的一部分。
结合附图从下面的描述中可以更加全面地理解本发明教导的前述和其他方面、实施例和特征。本发明的其他附加方面例如示例性实施方式的特征和/或有益效果将在下面的描述中显见,或通过根据本发明教导的具体实施方式的实践中得知。
附图不意在按比例绘制。在附图中,在各个图中示出的每个相同或近似相同的组成部分可以用相同的标号表示。为了清晰起见,在每个图中,并非每个组成部分均被标记。现在,将通过例子并参考附图来描述本发明的各个方面的实施例,其中:
图1是本发明的基于多源信息融合的移动电子商务推荐方法的流程图。
为了更了解本发明的技术内容,特举具体实施例并配合所附图式说明如下。
结合图1,本发明提出一种基于多源信息融合的移动电子商务推荐方法,所述方法包括:
S1:从多个信息源获取用户信息和对应的消费数据,对获取的用户信息和对应的消费数据进行预处理,以获取原始证据体,所述原始证据体被划分成若干类。
信息源包括云数据,各个购物平台的数据库,用户浏览的网页信息、视频信息,搜索引擎提供的用户数据,用户的聊天记录数据,存储在用户移动终端上的位置信息、语音、图片、视频等。
本发明提出,优选的,从所有可能获取到用户信息的信息源来获取信息,以全面了解消费者的需求,再根据需求推荐产品。尤其是位置信息,消费者的位置与他们的购物行为之间存在密切关系,为了使推荐系统发挥更大的作用,移动电子商务推荐系统可以选择整合用户位置信息,即,将多源信息融合方法引入到位置和历史行为信息的整合中。
获取方法包括以下几种:通过“昵称-手机号码”以提取用户特征数据以及用户的高频关注数据,通过“平台ID-手机号码”以从各个购物平台的数据库中挖掘用户的高频搜索数据与历史购买信息等等。
本发明针对移动电子商务中多源信息融合与决策推荐,使用位置信息、社交平台评论、以及产品信息和用户信息,有效增加推荐准确性。该方法弥补了以往推荐方法仅关注产品以及用户在消费平台的购物信息,不结合用户的其他信息的准确性缺陷。
进一步的实施例中,步骤S1中,对获取的用户信息和对应的消费数据进行预处理包括:
采用微格式对获取的用户信息和对应的消费数据进行表示、存储、集成和管理。
通过对获取的用户信息和对应的消费数据进行预处理,以使获取的数据格式统一,便于用户阅读和/或机器处理。
S2:基于所述原始证据体,通过径向基函数和神经网络算法,计算原始证据体中每类数据的证据权重,所述证据权重用于区分每类数据的推荐价值。
具体的,本发明通过采用一种改进的径向基神经网络以计算原始证据体中每类数据的证据权重。
其中,c为输出分类数目,x为神经元输入,y为神经元输出,σ
j为方差,j为输入个数。
S3:根据所述证据权重,运用D-S证据理论对原始证据体进行信息融合,以获得新的证据体。
为避免D-S理论的鲁棒性和一票否决性,基于D-S证据理论,引入证据可行度μ作为评判标准以修改原始证据体M,根据下述公式计算修改后的原始证据体M的基本概率分配值:
M=[m(A
1)m(A
2)…m(A
n)m(Θ)]
优选的,所述证据可行度μ的影响因子包括信息数据的时效性、稳定性和全面性中的一种 或者多种。
对原始证据体M完成修改之后,再根据基于局部冲突分配策略的合成规则,对修改后的原始证据体M进行融合,以获得新的证据体,采用下述公式计算融合后所得新证据体的基本概率分配值:
通过采用可行度来修改证据体、以及根据局部冲突分配合成规则这两个方面的结合,既能提高推荐的准确性,降低推荐系统响应时间,又能增加推荐覆盖率。
优选的,响应于每两个原始证据合成新的证据,对新的证据进行归一化处理,以便于后续处理。
S4:综合考虑信息的时效性,稳定性,全面性等众多因素,采用功率谱估计方法对所述新的证据体进行处理,以获取推荐决策。
具体的,根据用户需求和特征信息x(n)的自相关函数
代表任意两个不同用户在不同时刻的用户需求和特征信息x
N(n)与x
N(n+m)之间的相关程度,由该相关程度计算出功率谱密度P
xx(ω)=∑
mΦ
xx(m)e
-zωm,与相关程度为一对傅里叶变换。
基于前述方法,本发明还提及一种基于多源信息融合的移动电子商务推荐系统,所述系统包括以下几个模块:
1)用以从多个信息源获取用户信息和对应的消费数据,对获取的用户信息和对应的消费数据进行预处理,以获取原始证据体的模块,所述原始证据体被划分成若干类。
2)用以基于所述原始证据体,通过径向基函数和神经网络算法,计算原始证据体中每类数据的证据权重的模块,所述证据权重用于区分每类数据的推荐价值。
3)用以根据所述证据权重,运用D-S证据理论对原始证据体进行信息融合,以获得新的证据体的模块。
4)用以采用功率谱估计方法对所述新的证据体进行处理,以获取推荐决策的模块。
在本公开中参照附图来描述本发明的各方面,附图中示出了许多说明的实施例。本公开的实施例不必定义在包括本发明的所有方面。应当理解,上面介绍的多种构思和实施例,以及下面更加详细地描述的那些构思和实施方式可以以很多方式中任意一种来实施,这是因为本发明所公开的构思和实施例并不限于任何实施方式。另外,本发明公开的一些方面可以单独使用,或者与本发明公开的其他方面的任何适当组合来使用。
虽然本发明已以较佳实施例揭露如上,然其并非用以限定本发明。本发明所属技术领域中具有通常知识者,在不脱离本发明的精神和范围内,当可作各种的更动与润饰。因此,本发明的保护范围当视权利要求书所界定者为准。
Claims (9)
- 一种基于多源信息融合的移动电子商务推荐方法,其特征在于,所述方法包括:S1:从多个信息源获取用户信息和对应的消费数据,对获取的用户信息和对应的消费数据进行预处理,以获取原始证据体,所述原始证据体被划分成若干类;S2:基于所述原始证据体,通过径向基函数和神经网络算法,计算原始证据体中每类数据的证据权重,所述证据权重用于区分每类数据的推荐价值;S3:根据所述证据权重,运用D-S证据理论对原始证据体进行信息融合,以获得新的证据体;S4:采用功率谱估计方法对所述新的证据体进行处理,以获取推荐决策。
- 根据权利要求1所述的基于多源信息融合的移动电子商务推荐方法,其特征在于,步骤S1中,从多个信息源获取用户信息和对应的消费数据包括:通过“昵称-手机号码”以提取用户特征数据以及用户的高频关注数据,通过“平台ID-手机号码”以从各个购物平台的数据库中挖掘用户的高频搜索数据与历史购买信息。
- 根据权利要求1所述的基于多源信息融合的移动电子商务推荐方法,其特征在于,步骤S1中,对获取的用户信息和对应的消费数据进行预处理包括:采用微格式对获取的用户信息和对应的消费数据进行表示、存储、集成和管理。
- 根据权利要求1所述的基于多源信息融合的移动电子商务推荐方法,其特征在于,步骤S3中,根据所述证据权重,运用D-S证据理论对原始证据体进行信息融合,以获得新的证据体的方法包括:根据D-S证据理论,引入证据可行度μ作为评判标准以修改原始证据体M,根据下述公式计算修改后的原始证据体M的基本概率分配值:M=[m(A 1)m(A 2)…m(A n)m(Θ)]根据基于局部冲突分配策略的合成规则,对修改后的原始证据体M进行融合,以获得新的证据体,采用下述公式计算融合后所得新证据体的基本概率分配值:
- 根据权利要求5所述的基于多源信息融合的移动电子商务推荐方法,其特征在于,所述证据可行度μ的影响因子包括信息数据的时效性、稳定性和全面性中的一种或者多种。
- 根据权利要求1或者5所述的基于多源信息融合的移动电子商务推荐方法,其特征在于,所述方法还包括:响应于每两个原始证据合成新的证据,对新的证据进行归一化处理。
- 一种基于多源信息融合的移动电子商务推荐系统,其特征在于,所述系统包括:用以从多个信息源获取用户信息和对应的消费数据,对获取的用户信息和对应的消费数据进行预处理,以获取原始证据体的模块,所述原始证据体被划分成若干类;用以基于所述原始证据体,通过径向基函数和神经网络算法,计算原始证据体中每类数据的证据权重的模块,所述证据权重用于区分每类数据的推荐价值;用以根据所述证据权重,运用D-S证据理论对原始证据体进行信息融合,以获得新的证据体的模块;用以采用功率谱估计方法对所述新的证据体进行处理,以获取推荐决策的模块。
Applications Claiming Priority (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201910030517.8A CN109785064A (zh) | 2019-01-14 | 2019-01-14 | 一种基于多源信息融合的移动电子商务推荐方法和系统 |
CN201910030517.8 | 2019-01-14 |
Publications (1)
Publication Number | Publication Date |
---|---|
WO2020147265A1 true WO2020147265A1 (zh) | 2020-07-23 |
Family
ID=66500468
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
PCT/CN2019/092977 WO2020147265A1 (zh) | 2019-01-14 | 2019-06-26 | 一种基于多源信息融合的移动电子商务推荐方法和系统 |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN109785064A (zh) |
WO (1) | WO2020147265A1 (zh) |
Cited By (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113051249A (zh) * | 2021-03-22 | 2021-06-29 | 江苏杰瑞信息科技有限公司 | 一种基于多源异构大数据融合的云服务平台设计方法 |
CN113609360A (zh) * | 2021-08-19 | 2021-11-05 | 武汉东湖大数据交易中心股份有限公司 | 一种基于场景化多源数据融合分析的方法和系统 |
CN113792805A (zh) * | 2021-09-16 | 2021-12-14 | 河北工程大学 | 一种基于多源数据融合的水质评价方法 |
CN114757295A (zh) * | 2022-04-28 | 2022-07-15 | 浙江科技学院 | 基于云模型和证据理论的多传感器数据融合方法及应用 |
CN114997339A (zh) * | 2022-08-01 | 2022-09-02 | 白杨时代(北京)科技有限公司 | 一种多源目标智能决策方法及相关装置 |
CN115225301A (zh) * | 2021-04-21 | 2022-10-21 | 上海交通大学 | 基于d-s证据理论的混合入侵检测方法和系统 |
CN117150440A (zh) * | 2023-11-01 | 2023-12-01 | 人民法院信息技术服务中心 | 一种基于信息不确定性的多源信息融合方法及装置 |
Families Citing this family (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109785064A (zh) * | 2019-01-14 | 2019-05-21 | 南京信息工程大学 | 一种基于多源信息融合的移动电子商务推荐方法和系统 |
CN110348964A (zh) * | 2019-07-09 | 2019-10-18 | 葛晓滨 | 一种基于多感知的智慧化电子商务推荐方法 |
CN112800999B (zh) * | 2021-02-04 | 2023-12-01 | 大连大学 | 基于智联网的智能指控系统目标识别融合方法 |
CN112989207B (zh) * | 2021-04-27 | 2021-08-27 | 武汉卓尔数字传媒科技有限公司 | 一种信息推荐方法及装置、电子设备、存储介质 |
CN113988060B (zh) * | 2021-08-31 | 2024-09-17 | 中国电子科技集团公司第十五研究所 | 数据信息推荐方法、装置、终端设备和存储介质 |
Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1643920A (zh) * | 2002-03-19 | 2005-07-20 | 皇家飞利浦电子股份有限公司 | 使用多个推荐得分的推荐系统 |
CN103544539A (zh) * | 2013-10-12 | 2014-01-29 | 国家电网公司 | 一种基于人工神经网络和d-s证据理论的用户变化量预测方法 |
US20150178265A1 (en) * | 2013-12-20 | 2015-06-25 | Google Inc. | Content Recommendation System using a Neural Network Language Model |
CN106327240A (zh) * | 2016-08-11 | 2017-01-11 | 中国船舶重工集团公司第七0九研究所 | 一种基于gru神经网络的推荐方法和系统 |
CN107341447A (zh) * | 2017-06-13 | 2017-11-10 | 华南理工大学 | 一种基于深度卷积神经网络和证据k近邻的人脸核实方法 |
CN109146644A (zh) * | 2018-09-05 | 2019-01-04 | 广州小楠科技有限公司 | 一种电子商务系统 |
CN109785064A (zh) * | 2019-01-14 | 2019-05-21 | 南京信息工程大学 | 一种基于多源信息融合的移动电子商务推荐方法和系统 |
Family Cites Families (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN103064856B (zh) * | 2011-10-21 | 2016-03-30 | 中国移动通信集团重庆有限公司 | 一种基于信度网的资源推荐方法和装置 |
US10319022B2 (en) * | 2013-02-28 | 2019-06-11 | Lg Electronics Inc. | Apparatus and method for processing a multimedia commerce service |
US11769193B2 (en) * | 2016-02-11 | 2023-09-26 | Ebay Inc. | System and method for detecting visually similar items |
CN107909433A (zh) * | 2017-11-14 | 2018-04-13 | 重庆邮电大学 | 一种基于大数据移动电子商务的商品推荐方法 |
-
2019
- 2019-01-14 CN CN201910030517.8A patent/CN109785064A/zh active Pending
- 2019-06-26 WO PCT/CN2019/092977 patent/WO2020147265A1/zh active Application Filing
Patent Citations (7)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN1643920A (zh) * | 2002-03-19 | 2005-07-20 | 皇家飞利浦电子股份有限公司 | 使用多个推荐得分的推荐系统 |
CN103544539A (zh) * | 2013-10-12 | 2014-01-29 | 国家电网公司 | 一种基于人工神经网络和d-s证据理论的用户变化量预测方法 |
US20150178265A1 (en) * | 2013-12-20 | 2015-06-25 | Google Inc. | Content Recommendation System using a Neural Network Language Model |
CN106327240A (zh) * | 2016-08-11 | 2017-01-11 | 中国船舶重工集团公司第七0九研究所 | 一种基于gru神经网络的推荐方法和系统 |
CN107341447A (zh) * | 2017-06-13 | 2017-11-10 | 华南理工大学 | 一种基于深度卷积神经网络和证据k近邻的人脸核实方法 |
CN109146644A (zh) * | 2018-09-05 | 2019-01-04 | 广州小楠科技有限公司 | 一种电子商务系统 |
CN109785064A (zh) * | 2019-01-14 | 2019-05-21 | 南京信息工程大学 | 一种基于多源信息融合的移动电子商务推荐方法和系统 |
Cited By (11)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN113051249A (zh) * | 2021-03-22 | 2021-06-29 | 江苏杰瑞信息科技有限公司 | 一种基于多源异构大数据融合的云服务平台设计方法 |
CN115225301A (zh) * | 2021-04-21 | 2022-10-21 | 上海交通大学 | 基于d-s证据理论的混合入侵检测方法和系统 |
CN115225301B (zh) * | 2021-04-21 | 2023-11-21 | 上海交通大学 | 基于d-s证据理论的混合入侵检测方法和系统 |
CN113609360A (zh) * | 2021-08-19 | 2021-11-05 | 武汉东湖大数据交易中心股份有限公司 | 一种基于场景化多源数据融合分析的方法和系统 |
CN113792805A (zh) * | 2021-09-16 | 2021-12-14 | 河北工程大学 | 一种基于多源数据融合的水质评价方法 |
CN113792805B (zh) * | 2021-09-16 | 2024-05-10 | 河北工程大学 | 一种基于多源数据融合的水质评价方法 |
CN114757295A (zh) * | 2022-04-28 | 2022-07-15 | 浙江科技学院 | 基于云模型和证据理论的多传感器数据融合方法及应用 |
CN114757295B (zh) * | 2022-04-28 | 2024-04-02 | 浙江科技学院 | 基于云模型和证据理论的多传感器数据融合方法及应用 |
CN114997339A (zh) * | 2022-08-01 | 2022-09-02 | 白杨时代(北京)科技有限公司 | 一种多源目标智能决策方法及相关装置 |
CN117150440A (zh) * | 2023-11-01 | 2023-12-01 | 人民法院信息技术服务中心 | 一种基于信息不确定性的多源信息融合方法及装置 |
CN117150440B (zh) * | 2023-11-01 | 2024-02-06 | 人民法院信息技术服务中心 | 一种基于信息不确定性的多源信息融合方法及装置 |
Also Published As
Publication number | Publication date |
---|---|
CN109785064A (zh) | 2019-05-21 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
WO2020147265A1 (zh) | 一种基于多源信息融合的移动电子商务推荐方法和系统 | |
US8832091B1 (en) | Graph-based semantic analysis of items | |
Kumar et al. | Systematic Literature Review on Opinion Mining of Big Data for Government Intelligence. | |
WO2023065211A1 (zh) | 一种信息获取方法以及装置 | |
Yadav et al. | Dealing with Pure New User Cold‐Start Problem in Recommendation System Based on Linked Open Data and Social Network Features | |
Margaris et al. | An adaptive social network-aware collaborative filtering algorithm for improved rating prediction accuracy | |
TWI705411B (zh) | 社交業務特徵用戶的識別方法和裝置 | |
Xie et al. | Data mining and application of social e-commerce users based on big data of internet of things | |
Li et al. | A novel time-aware hybrid recommendation scheme combining user feedback and collaborative filtering | |
Karthik et al. | A recommendation system for online purchase using feature and product ranking | |
Li et al. | Social recommendation based on trust and influence in SNS environments | |
Cao et al. | Web API recommendation via combining graph attention representation and deep factorization machines quality prediction | |
Mirtalaie et al. | Sentiment analysis of specific product’s features using product tree for application in new product development | |
He et al. | Online spam review detection: A survey of literature | |
Latrech et al. | CoDFi-DL: a hybrid recommender system combining enhanced collaborative and demographic filtering based on deep learning | |
Sun | Music Individualization Recommendation System Based on Big Data Analysis | |
Chafale et al. | Sentiment analysis on product reviews using Plutchik’s wheel of emotions with fuzzy logic | |
Pasricha et al. | A new approach for book recommendation using opinion leader mining | |
Hidayati et al. | The Influence of User Profile and Post Metadata on the Popularity of Image-Based Social Media: A Data Perspective | |
Deng et al. | A Trust-aware Neural Collaborative Filtering for Elearning Recommendation. | |
Gupta et al. | Scalable online product recommendation engine based on implicit feature extraction domain | |
Hanafi et al. | Word Sequential Using Deep LSTM and Matrix Factorization to Handle Rating Sparse Data for E‐Commerce Recommender System | |
Zhu et al. | Intelligent product redesign strategy with ontology-based fine-grained sentiment analysis | |
Sun et al. | Enabling 5G: sentimental image dominant graph topic model for cross-modality topic detection | |
Wang | Application of E-Commerce Recommendation Algorithm in Consumer Preference Prediction |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
121 | Ep: the epo has been informed by wipo that ep was designated in this application |
Ref document number: 19910488 Country of ref document: EP Kind code of ref document: A1 |
|
NENP | Non-entry into the national phase |
Ref country code: DE |
|
122 | Ep: pct application non-entry in european phase |
Ref document number: 19910488 Country of ref document: EP Kind code of ref document: A1 |