In the purchase of products, a warehouse is an integral part between producers and customers, where the order picking is one of several operations involved. This chapter highlights the prioritization of the selection of the elements that make up an order based on the influence of the customer's personality on their preferences in order to choose the alternatives of the order. For this purpose, two approaches of personality theories are integrated in order to model the influence of personality as a factor that influences the parameter values of preference models based on outranking relations. The application case deals with an online supermarket using an intelligent virtual agent as the assistant that receives the order, who emulates the personality and preferences of the customer, selecting and delivering the best order. This chapter will emulate the behavior of decision makers, showing the impact of personality on preferences and will analyze its range of applications in problems related to the order picking.
TopIntroduction
The virtual world of the Internet has become a very important platform to optimize, promote and facilitate the ways of acquiring products of all kinds, from everyday poducts to especially those that for some reason are not commercialized in the locality, for example, imported or manufactured products in a specific region, such as clothing, medical supplies, food, electronic devices, etc. In this type of transaction, Warehouse is an integral part that forms a link between producers and customers, guaranteeing the delivery of the goods under negotiation. The operations involved in the storage can be divided into several functions, such as Receiving, Transfer, Put Away, Order-Picking and Shipping (D. M. Lambert, Stock, & Ellram, 1998; Koster, Le-Duc, & Roodbergen, 2007; Horvat, 2012), this research work focuses on the selection of the elements that make up an order (Order-Picking).
Order-Picking consists in taking and collecting the quantities of items in the warehouse to satisfy the customer's demand (D. M. Lambert et al., 1998; Koster et al., 2007; Horvat, 2012). This collection process is prone to a series of problems, such as not placing the correct items requested in the order in the warehouse, collection errors by personnel, etc. (Bellwood, 2015). The priority factor of the order is also surrounded by various circumstances to decide which order should be assigned more quickly than others. Whatever the reason, it means that some orders have higher priority than others and, therefore, this influences the selection strategy (Saleswarp, 2018).
The priority of the order, the shipping requirements, the date / age of the order, the location of the collector and the availability of the inventory are also factors that influence the items selection of orders. However, there is also possible prioritize the items selection that make up the order based on the customer's preferences. These preferences imply criteria of importance for the selection, such as brand, price, content, presentation, weight, units, etc. In case of shortage in the inventory of any of the requested products, the collector can select a similar missing element taking into account the customer preference's criteria.
This last case of prioritization is the one that will be addressed in this work; where the selection of the items that make up an order is based on a preferences model and the influence of the customer's personality profile on it in order to choose an alternative item.
This chapter presents a basic case of study of online purchases, where an Intelligent Virtual Agent (IVA) as a virtual assistant interacts with the customer in order to select the requested warehouse elements, prioritizing the selection based on the preferences and personality of the customer.
In order to facilitate decision making and thus choose the most convenient solution, the agent architecture uses the NOSGA-II preferences model, which, will be influenced by the personality of the customer. NOSGA-II (Fernández, López, López, & Coello-Coello, 2011) is a Non-Outranking Sorting Genetic Algorithm that takes into account the preferences of a decision maker. This model uses parameters such as: the weights (indicate the priority of the criteria involved in each product), and thresholds of preference (indifference, preveto and veto), which are used to compare the different alternatives. These parameters of preference are those who will receive the influence of the personality of the decision maker, so this work proposes a strategy that conveniently fits these parameters of preference through a model of personality influences built from two personality models that are the most recurrent in the literature: KTS (Keirsey Temperament Sorted) and OCEAN (Five Factor Model), whose designs are based on two complementary personality approaches. Therefore the personality influence model proposed is the result of an analysis of KTS and OCEAN from which a set of profiles that aim to describe the behavior of an individual in a decision context is derived.