Search results

A good decision support system for credit scoring enables telecom operators to measure the subscribers' creditworthiness in a fine-grained manner. This paper aims to propose a robust credit scoring system by leveraging latent information embedded in the telecom subscriber relation network based on multi-source data sources, including telecom inner data, online app usage, and offline consumption footprint.

Rooting from network science, the relation network model and singular value decomposition are integrated to infer different subscriber subgroups. Employing the results of network inference, the paper proposed a network-aware credit scoring system to predict the continuous credit scores by implementing several state-of-art techniques, i.e. multivariate linear regression, random forest regression, support vector regression, multilayer perceptron, and a deep learning algorithm. The authors use a data set consisting of 926 users of a Chinese major telecom operator within one month of 2018 to verify the proposed approach.

The distribution of telecom subscriber relation network follows a power-law function instead of the Gaussian function previously thought. This network-aware inference divides the subscriber population into a connected subgroup and a discrete subgroup. Besides, the findings demonstrate that the network-aware decision support system achieves better and more accurate prediction performance. In particular, the results show that our approach considering stochastic equivalence reveals that the forecasting error of the connected-subgroup model is significantly reduced by 7.89–25.64% as compared to the benchmark. Deep learning performs the best which might indicate that a non-linear relationship exists between telecom subscribers' credit scores and their multi-channel behaviours.

This paper contributes to the existing literature on business intelligence analytics and continuous credit scoring by incorporating latent information of the relation network and external information from multi-source data (e.g. online app usage and offline consumption footprint). Also, the authors have proposed a power-law distribution-based network-aware decision support system to reinforce the prediction performance of individual telecom subscribers' credit scoring for the telecom marketing domain.

The objective of this paper is to propose a consumer-behavior-based intelligence (CBBI) model to identify market structure so as to monitor product competition. Competitive intelligence extracted from Chinese e-business clickstream data is exploited to examine the relevance of consumers' heterogeneous behavioral feedback, namely, click, tag-into-favorite, time-of-browsing, add-into-cart, and remove-from-cart, to visualize the competitive product market structure and to predict product-level sales.

Our proposed CBBI model consists of visualization and prediction, which explore e-business clickstream data. We conduct the visualization and segmentation of market structure in the form of a perceptual map by employing K-means clustering algorithm and multidimensional scaling technique. Concurrently, we developed an updated Bayesian linear regression (BLR) to predict product-level sales by considering consumers' heterogeneous feedback. Our updated BLR specifically integrated the estimated knowledge of the previous periods to verify whether product sales are period-dependent due to the consumer memory effect in e-commerce, improving the conventional BLR of diffuse prior distribution setup in terms of mean absolute error (MAE) and root mean squared error (RMSE).

Considering the performance of consumers' heterogeneous actions, the present research visualized three different segments of the competitive market structure in a perceptual map, and its horizontal axis is shown as a signal of the ascending trend of product sales. The previous five-day period was ascertained to be the best size of a time window for the consumer memory effect on product sales prediction. This hypothesis is supported by the concept that product sales are period-dependent. The results of the proposed updated BLR indicate that consumer tag-into-favorite, add-into-cart, and remove-from-cart feedback have positive impacts on product-level sales while click and time-of-browsing have the opposite effect.

While the identified competitive product market structure elaborates consumer heterogeneous feedback toward alternative product choices, this paper contributes by extending those homogeneous consumer preferences-related marketing studies. The perceptual map's configuration in respect to period-dependent product sales facilitates the effective inclusion of consumer behavior application in product sales prediction research in e-commerce. This paper helps sellers and retailers better comprehend the impacts of heterogeneous feedback and the consumer memory effect on the degree of competition in the form of product sales. The research results also offer a managerial implication about shaping the competitive edge by conducting different product management strategies in e-commerce platforms.

This paper aims to investigate the effect and mechanism of O atom single doping, Ce and O atoms co-doping on the interfacial microscopic behavior of brazed Ni-Cr/diamond.

Using first-principles calculations, the embedding energy, work of separation, interfacial energy and electronic structures of Ni-Cr-O/diamond and Ni-Cr-O-Ce/diamond interface models were calculated. Then, the effect of Ce and O co-doping was experimentally verified through brazed diamond with CeO₂-added Ni-Cr filler alloy.

The results show that O single-doping reduces the interfacial bonding strength between Ni-Cr filler alloy and diamond but enhances its interfacial stability to some extent. However, the Ce and O co-doping simultaneously enhances the interfacial bonding strength and stability between Ni-Cr filler alloy and diamond. The in-situ formed Ce-O oxide at interface impedes the direct contact between diamond and Ni-Cr filler alloy, which weakens the catalytic effect of Ni element on diamond graphitization. It is experimentally found that the fine rod-shaped Cr₃C₂ and Cr₇C₃ carbides are generated on diamond surface brazed with CeO₂-added Ni-Cr filler alloy. After grinding, the brazed diamond grits, brazed with CeO₂-added Ni-Cr filler alloy, present few fracture and the percentage of intact diamond reaches 67.8%. Compared to pure Ni-Cr filler alloy, the brazed diamond with CeO₂-added Ni-Cr filler alloy exhibit the better wear resistance and the slighter thermal damage.

Using first-principles calculations, the effect of Ce and O atoms co-doping on the brazed diamond with Ni-Cr filler alloy is investigated, and the calculation results are verified experimentally. Through the first-principles calculations, the interface behavior and reaction mechanism between diamond and filler alloy can be well disclosed, and the composition of filler alloy can be optimized, which will be beneficial for synergistically realizing the enhanced interface bonding and reduced thermal damage of brazed diamond.