Historical Purchase Data Research Articles

MHANER: A Multi-source Heterogeneous Graph Attention Network for Explainable Recommendation in Online Games

Recommender system helps address information overload problem and satisfy consumers’ personalized requirement in many applications such as e-commerce, social networks and in-game store. However, existing approaches mainly focus on improving the accuracy of recommendation tasks, but usually ignore to improve the interpretability of recommendation, which is still a challenging and crucial task, especially for some complicated scenarios such as large-scale online games. Some few previous attempts on explainable recommendation mostly depend on a large amount of a priori knowledge or user-provided review corpus, which is labor-consuming as well as often suffers from data deficiency. To relieve this issue, we propose a Multi-source Heterogeneous Graph Attention Network for Explainable Recommendation (MHANER) for the case without enough a priori knowledge or corpus of user comments. Specifically, MHANER employs the attention mechanism to model players’ preference to in-game store items as the support for the explanation of recommendation. Then a graph neural network based method is designed to model players’ multi-source heterogeneous information, including the players’ historical behavior data, historical purchase data, and attributes of the player-controlled character, which is leveraged to recommend possible items for players to buy. Finally, the multi-level subgraph pattern mining is adopted to combine the characteristics of a recommendation list to generate corresponding explanations of items. Extensive experiments on three real-world datasets, two collected from JD and one from NetEase game, demonstrate that the proposed model MHANER outperforms state-of-the-art baselines. Moreover, the generated explanations are verified by human encoding comprised of hard-core game players and endorsed by experts from game developers.

PREDICTION OF CUSTOMER CHURN USING ML

No business can thrive without it’s customers. On the flip side, customers leaving the business is a nightmare that every business owner dreads!In fact, one of the key metrics to measure a business’ success is by measuring its customer churn rate - the lower the churn, the more loved the company is.Typically, every user of a product or a service is assigned a prediction value that estimates their state of churn at any given time. This value may be based on multiple factors such as the user’s demographic, their browsing behavior and historical purchase data, among other details.This value factors in unique and proprietary predictions of how long a user will remain a customer and is updated every day for all users who have purchased at least one of the products/services. The values assigned are between 1 and 5

ALBERT-BPF: a book purchase forecast model for university library by using ALBERT for text feature extraction

PurposeThe purpose of this research is to achieve automatic and accurate book purchase forecasts for the university libraries and improve efficiency of manual book purchase.Design/methodology/approachThe authors presented a Book Purchase Forecast model with A Lite BERT(ALBERT-BPF) to achieve their goals. First, the authors process all the book data to unify format of books' features, such as ISBN, title, authors, brief introduction and so on. Second, they exploit the book order data to label all books supplied by booksellers with “purchased” or “non-purchased”. The labelled data will be used for model training. Last, the authors regard the book purchase task as a text classification problem and present a model named ALBERT-BPF, which applies ALBERT to extract text features of books and BPF classification layer to forecast purchased books, to solve the problem.FindingsThe application of deep learning in book purchase task is effective. The data the authors exploited are the historical book purchase data from their university library. The authors’ experiments on the data show that ALBERT-BPF can seek out the books that need to be purchased with an accuracy of over 82%. And the highest accuracy reached is 88.06%. These indicate that the deep learning model is sufficient to assist the traditional manual book purchase way.Originality/valueThis research applies ALBERT, which is based on the latest Natural Language Processing (NLP) architecture Transformer, to library book purchase task.

A novel temporal recommendation method based on user query topic evolution

Personalized product recommendation provides online customers with options in line with consumers’ interests and preferences to assist users in making purchasing decisions. Previous personalized recommendation methods identify user preferences by analyzing historical purchase data. As the user preferences reflected in the user–item matrix take only two values: like or dislike, the binary purchase data-based method is too constrained to capture individual diverse opinions and user preferences evolution. Despite the fact that online search queries explicitly express consumers’ content preferences for product attributes, very little research has focused on inferring user preferences from online queries. To this end, this study proposes a novel temporal recommendation method based on estimating consumers’ dynamic preferences from their search queries. Our approach involves: first, proposing a temporal topic model that utilizes a Markov process to predict users’ potential interests, and then integrating customer preferences and product attribute features into a unified product ranking model to obtain personalized and diversified product lists. Empirical analysis shows that the proposed approach can better capture the temporal dynamics of customer preferences drift and can more accurately predict consumer purchase intentions. Furthermore, products recommended by the proposed approach not only better match individual preferences, but also cover more diverse content topics.

New fuzzy RFMD model. A method to identify new customer profiles due to an increase in online activity. Case study of a retail business

Gaining customer loyalty has become one of the main objectives of all companies. Retailers, especially the online ones, have the advantage of knowing their customers’ historical purchase data, which provides them with an understanding of the customers’ buying patterns. A widely-used tool in strategic marketing and customer loyalty is segmentation based on the traditional Recency, Frequency and Monetary (RFM) model. Subsequently, the fuzzy RFM model proved to be an improvement on the traditional RFM model. There has been a change in the retail customer profile, with the growth of a new cluster, the “One-Shot Customer”, new customers that buy from a retailer just once and never come back. In response to this change, the fuzzy RFM model has been modified to include a new dimension capturing Length or Duration. This study presents the new fuzzy RFMD model (Recency, Frequency, Monetary and Duration model), which can be used to better identify that new, large group of customers. The paper also provides a case study based on an e-commerce clothing retailer. Its customer database was segmented using the k-means algorithm and the Isolation Forest algorithm was applied to identify and correctly treat possible anomalies. The Customer Lifetime Value and the weights for the RFMD attributes were calculated by applying the Analytic Hierarchy Process (AHP) model. Results reveal the improvement that the weighted fuzzy RFMD model offers to retailers, enabling them to detect the One-Shot Customers and thus optimize their strategic marketing plans.

Choosing quantities impacts individuals choice, rationality, and willingness to pay estimates

AbstractConventional choice experiments ask respondents to choose their most preferred item, but in reality, consumers select both an item and quantity. Using a split sample design focused on U.S. meat demand, we find that flexibility in consumer quantity choice provides similar choice rationality and consumer segmentation as the conventional restricted quantity framework. Improvements include generating choice frequencies consistent with historical purchase data and willingness‐to‐pay estimates more comparable to observed retail price data. Additional benefits include the ability to understand consumer preferences for variety and habit formation and generated data that is compatible with traditional demand estimation, all of which are not possible under the conventional restricted quantity framework.

A Shapley Value based Approach to Market Basket Analysis

Market Basket Analysis is an interesting concept wherein a set of historical purchase data is studied extensively and data-mining techniques are applied to predict a user’s purchase behavior. In the resent times, every e-commerce giant, like Amazon, Flip-kart etc. are trying to increase their sales by this technique. It has been found in study that when a user is purchasing an item then he/she might be interested in purchasing other items too. So, by analyzing this purchase behavior if we can recommend other products at the time of purchasing then that could increase the sale and profit as well. Analyzing the correlation among the products by studying the purchase pattern and then finding the associated products is the main objective of Market Basket Analysis. Several algorithms have been developed to find the association among the products. But most of the traditional algorithms are based on support value of the product. They do not consider another important factor i.e. the marginal contribution of the product. Considering this into account in this paper we are going to propose a Shapley Value based game theoretic approach for market basket analysis. Shapley Value is a well-known solution concept in cooperative game theory. It gives the measure of marginal contribution of a player in a cooperative game.

CFSH: Factorizing sequential and historical purchase data for basket recommendation.

To predict what products customers will buy in next transaction is an important task. Existing work in next-basket prediction can be summarized into two paradigms. One is the item-centric paradigm, where sequential patterns are mined from customers’ transactional data and leveraged for prediction. However, these approaches usually suffer from the data sparseness problem. The other is the user-centric paradigm, where collaborative filtering techniques have been applied on customers’ historical data. However, these methods ignore the sequential behaviors of customers which is often crucial for next-basket prediction. In this paper, we introduce a hybrid method, namely the Co-Factorization model over Sequential and Historical purchase data (CFSH for short) for next-basket recommendation. Compared with existing methods, our approach conveys the following merits: 1) By mining global sequential patterns, we can avoid the sparseness problem in traditional item-centric methods; 2) By factorizing product-product and customer-product matrices simultaneously, we can fully exploit both sequential and historical behaviors to learn customer and product representations better; 3) By using a hybrid recommendation method, we can achieve better performance in next-basket prediction. Experimental results on three real-world purchase datasets demonstrated the effectiveness of our approach as compared with the state-of-the-art methods.

Product Selection Problem

It is often crucial for manufacturers to decide what products to produce so that they can increase their market share in an increasingly fierce market. To decide which products to produce, manufacturers need to analyze the consumers’ requirements and how consumers make their purchase decisions so that the new products will be competitive in the market. In this paper, we first present a general distance-based product adoption model to capture consumers’ purchase behavior. Using this model, various distance metrics can be used to describe different real life purchase behavior. We then provide a learning algorithm to decide which set of distance metrics one should use when we are given some accessible historical purchase data. Based on the product adoption model, we formalize the k most marketable products (or k- MMP ) selection problem and formally prove that the problem is NP-hard . To tackle this problem, we propose an efficient greedy-based approximation algorithm with a provable solution guarantee. Using submodularity analysis, we prove that our approximation algorithm can achieve at least 63% of the optimal solution. We apply our algorithm on both synthetic datasets and real-world datasets (TripAdvisor.com), and show that our algorithm can easily achieve five or more orders of speedup over the exhaustive search and achieve about 96% of the optimal solution on average. Our experiments also demonstrate the robustness of our distance metric learning method, and illustrate how one can adopt it to improve the accuracy of product selection.

Product/service variety strategy considering mixed distribution of human lifestyles

Managing product and service variety for a retail store is a key task to satisfy various customer needs given some constraints. Nevertheless, it is difficult to ascertain the distribution of customer demands because customers have different lifestyles and valuation bases. This paper presents a proposal of an analytical method for product and service variety using historical purchase data and a lifestyle survey. It introduces a study example of a supermarket chain in which customers’ purchase behaviors can be elucidated more clearly by categorizing customers into six lifestyle categories. Results illustrate how customer data in services are useful for manufacturing.

A Data Mining-Based Response Model for Target Selection in Direct Marketing

Identifying customers who are more likely to respond to new product offers is an important issue in direct marketing. In direct marketing, data mining has been used extensively to identify potential customers for a new product (target selection). Using historical purchase data, a predictive response model with data mining techniques was developed to predict a probability that a customer in Ebedi Microfinance bank will respond to a promotion or an offer. To achieve this purpose, a predictive response model using customers' historical purchase data was built with data mining techniques. The data were stored in a data warehouse to serve as management decision support system. The response model was built from customers' historic purchases and demographic dataset.

Constructing response model using ensemble based on feature subset selection

In building a response model, determining the inputs to the model has been an important issue because of the complexities of the marketing problem and limitations of mental models for decision-making. It is common that the customers' historical purchase data contains many irrelevant or redundant features thus result in bad model performance. Furthermore, single complex models based on feature subset selection may not always report improved performance largely because of overfitting and instability. Ensemble is a widely adopted mechanism for alleviating such problems. In this paper, we propose an ensemble creation method based on GA based wrapper feature subset selection mechanism. Through experimental studies on DMEF4 data set, we found that the proposed method has, at least, two distinct advantages over other models: first, the ability to account for the important inputs to the response model; second, improved prediction accuracy and stability.

Response models based on bagging neural networks

Identifying customers who are likely to respond to a product offering is an important issue in direct marketing.Response models are typically built from historical purchase data. A popular method of choice, logistic regression, is easy to understand and build, but limited in that the model is linear in parameters. Neural networks are nonlinear and have been found to improve predictive accuracies for a variety of business applications. Neural networks have not always demonstrated clear supremacy over traditional statistics competitors, largely because of over-fitting and instability. Combining multiple networksmay alleviate these problems. A systematic method of combining neural networks is proposed, namely bagging or bootstrap aggregating, whereby overfitted multiple neural networks are trained with bootstrap replicas of the original data set and then averaged. We built response models using a publicly available DMEF data set with three methods: bagging neural networks, single neural networks, and conventional logistic regression. The proposed method not only improved but also stabilized the prediction accuracies over the other two.