Fashion E-commerce Research Articles

A Semi-Supervised Learning Approach for Automatic Detection and Fashion Product Category Prediction with Small Training Dataset Using FC-YOLOv4

Over the past few decades, research on object detection has developed rapidly, one of which can be seen in the fashion industry. Fast and accurate detection of an E-commerce fashion product is crucial to choosing the appropriate category. Nowadays, both new and second-hand clothing is provided by E-commerce sites for purchase. Therefore, when categorizing fashion clothing, it is essential to categorize it precisely, regardless of the cluttered background. We present recently acquired tiny product images with various resolutions, sizes, and positions datasets from the Shopee E-commerce (Thailand) website. This paper also proposes the Fashion Category—You Only Look Once version 4 model called FC-YOLOv4 for detecting multiclass fashion products. We used the semi-supervised learning approach to reduce image labeling time, and the number of resulting images is then increased through image augmentation. This approach results in reasonable Average Precision (AP), Mean Average Precision (mAP), True or False Positive (TP/FP), Recall, Intersection over Union (IoU), and reliable object detection. According to experimental findings, our model increases the mAP by 0.07 percent and 40.2 percent increment compared to the original YOLOv4 and YOLOv3. Experimental findings from our FC-YOLOv4 model demonstrate that it can effectively provide accurate fashion category detection for properly captured and clutter images compared to the YOLOv4 and YOLOv3 models.

A Tale of Color Variants: Representation and Self-Supervised Learning in Fashion E-commerce

In this paper, we address a crucial problem in fashion e-commerce (with respect to customer experience, as well as revenue): color variants identification, i.e., identifying fashion products that match exactly in their design (or style), but only to differ in their color. We propose a generic framework, that leverages deep visual Representation Learning at its heart, to address this problem for our fashion e-commerce platform. Our framework could be trained with supervisory signals in the form of triplets, that are obtained manually. However, it is infeasible to obtain manual annotations for the entire huge collection of data usually present in fashion e-commerce platforms, such as ours, while capturing all the difficult corner cases. But, to our rescue, interestingly we observed that this crucial problem in fashion e-commerce could also be solved by simple color jitter based image augmentation, that recently became widely popular in the contrastive Self-Supervised Learning (SSL) literature, that seeks to learn visual representations without using manual labels. This naturally led to a question in our mind: Could we leverage SSL in our use-case, and still obtain comparable performance to our supervised framework? The answer is, Yes! because, color variant fashion objects are nothing but manifestations of a style, in different colors, and a model trained to be invariant to the color (with, or without supervision), should be able to recognize this! This is what the paper further demonstrates, both qualitatively, and quantitatively, while evaluating a couple of state-of-the-art SSL techniques, and also proposing a novel method.

Image-based virtual try-on system with clothing extraction module that adapts to any posture

In this study, we developed a virtual try-on system that automatically creates a clothing model from a person’s image, and thus a clothing model is not required as an input. Recently, the scale of the E-commerce market in the fashion field has expanded. When purchasing clothes on a fashion E-commerce site, a user can browse and buy various brands at once without visiting a physical store. However, users cannot pick up the actual clothes and try them on. Therefore, the user may not be satisfied with the design and size after receiving clothing bought on-line. Virtual try-on systems have been developed in various studies that allow users to virtually try on clothes using a computer. In particular, many image-based virtual try-on systems have been reported that transition the clothing model’s features to the clothing area on the target person. However, this method has the following problems: (1) it is necessary to prepare a clothing model in advance and (2) the posture for changing clothes is restricted. Thus, we developed a virtual try-on system that does not require a clothing model as an input, thereby enabling the automatic creation of a clothing model from a person’s image by incorporating a clothing extraction module into the virtual try-on system. In addition, we improved the virtual try-on system so it can adapt to any posture and evaluated the improvements in simulations.

A design in system architecture based on mobile cloud computing for a virtual try-on solution

Cloud computing is an emerging technology in this digital century. It provides an excellent but low-cost source for database storage, computing power, applications, and services through an internet-delivered cloud platform. Thanks to the cost savings in investing and maintaining physical data centers, as well as the stability of Quality of Service (QoS), it has no restrictions on company size or sector for enterprises to shift their operations to a cloud platform. The fashion industry, particularly the fashion e-commerce sector, is a case study in leveraging the cloud platform via a technology called “Virtual try-on” (VTO). VTO solution allows fashion brands to increase the shopping experience, however, requires installing and maintaining a bulky system for implementation. There are different methods and approaches to design architectures using cloud computing, however, there have not been many studies addressing tasks related to machine learning that uses the high Graphics Processing Unit (GPU) encountered in VTO solutions. To design a scheduler that could optimize the system performance while lowering operational expenses in VTO solutions, this research proposes a system to (1) handle synchronous model and asynchronous model separately and clearly, (2) perform multi-layered task processing architecture by hashing task ID and implementing a queue management system. This method would satisfy three major requirements: (1) Avoid complex hardware requirements for users, (2) Ensure the system stability and the ease of horizontal and vertical extension, and (3) Protect user information privacy.

Fine-Grained Identification of Clothing Apparels

Abstract: The rapid use of smartphones and tablet computers, search is now not just limited to text,but moved to other modalities such as voice and image. Extracting and matching this attributes still remains a daunting task due to high deformability and variability of clothing items. Visual analysis of clothings is a topic that has received attention due to tremendous growth of e-commerce fashion stores. Analyzing fashion attributes is also crucial in the fashion design process. This paper addresses the solution of recognition of clothes and fashion related attributes related to it using better image segmentation RCNN based algorithms. Keywords: Computer Vision, Fine grained identification, Clothing apparel detection, Convolutional Neural Network, Mask RCNN, Detectron2

Customer models for artificial intelligence-based decision support in fashion online retail supply chains

Fashion is a global, multi-trillion dollar industry devoted to producing and selling clothing, footwear, and accessories to individuals or groups of people. Its sheer numbers, together with social and environmental sustainability concerns, and the move towards digitalization of customer-centric operations, make the fashion business a prime target for Decision Support Systems (DSSs). On the other hand, decision support in fashion retail is particularly problematic and embraces all major supply chain domains. Decisions in an online fashion retail supply chain (FRSC) are highly dependent on time-varying customers' preferences and product availability, often leading to a combinatorial explosion. To address such a problem, DSSs could greatly benefit from high-quality information stored in customer models (CMs), constructed by using Artificial Intelligence techniques, allowing informed decisions on how to personalize (adapt) to match the customer's needs and preferences. Combinations of CMs with recommender systems (RSs) have been increasingly utilized in fashion e-commerce to provide personalized product recommendations. Nevertheless, works on enhancing CMs for e-commerce or other decision-making chain domains are scanty. This paper offers a systematic review of the literature on fashion CMs with applications to decision-making in FRSCs, mining topics for a research agenda. Research on the theme is relevant and urgent for the fashion business, which is still in its infancy. Work on the agenda topics could benefit distinct fashion stakeholders, not just customers, and produce well-grounded decision-making in varied FRSC contexts and dynamics.

Sizing up the body: Virtual fit platforms in fashion e-commerce

Dress is an embodied experience which is dematerialized online. In a fashion e-commerce website, clothes cannot be touched, nor worn prior to purchase and delivery; this engenders issues of fit and thus, returns. To solve this issue, fashion companies are turning to size recommendation and virtual fit service platforms. Simply put, virtual fit systems algorithmically match customer body data to fashion items which are potentially the right size and fit. This process aims to create value for all parties involved: for brands, by improving customer satisfaction and reducing returns; for customers, by facilitating choices; and for platform providers, by the sale of services and tools. However, as research in online platforms in other fields suggests (: 197; ), virtual fit services are driven by mechanisms of datafication, curation and commodification of fashion consumers’ bodily data – which in turn raise issues related to privacy and inclusivity. To the best of the authors’ knowledge, virtual fit platforms and their effects on the datafication of dress embodiment have heretofore not been discussed in fashion studies literature. This article spotlights the growing phenomena, opening avenues for further research in the field.

Search engine optimization: The long-term strategy of keyword choice

In digital marketing planning, keyword choice for search engine positioning is crucial for customer attraction, and the associated investment is an increasingly relevant part of the marketing budget. To what extent is it possible to optimize over time the resources allocated to positioning a brand on search engines? Often keyword costs are understood as monetary costs and the literature on keyword auctions suggests convergence towards a long-run equilibrium expenditure path. However, search engine optimization techniques, associated with implicit costs, have mainly focused on the short run. This paper suggests drawing long-term inferences from estimated cost per click, as the economic cost of organic keywords.Web analytics software is used to gather these data for leading fashion e-commerce. A time series test of transition dynamics identifies convergence of economic keyword costs between branded and generic keywords or catch-up, as an early state of convergence, for almost all considered e-commerce.

Fashion involvement, opinion-seeking and product variety as stimulators for fashion e-commerce: an investigated model based on S-O-R model

PurposeThis study's aim was to investigate the stimulators of fashion e-consumers within e-commerce environments. The study proposed a framework utilizing stimulus–organism–response (S-O-R) theory to suggest that fashion involvement and opinion-seeking would act as sociopsychological stimuli, while product variety as an objective stimulus. Perceived quality is proposed as an organism, moderated by perceived price. Consumer buying behavior within e-commerce environments presents the framework's response. The study looked at variables with deeper insights into Malaysian fashion consumers.Design/methodology/approachA quantitative method was used to assess the significance of relationships within the proposed model. Partial least squares structural equation modeling technique was implemented to assess the framework's relationships with a sample size of 374.FindingsResults indicate that fashion involvement is significantly associated as a sociopsychological stimulus, with product variety being an objective stimulus for Malaysian fashion e-consumers. Perceived quality is significantly represented as an organism through the framework, and buying behavior is the latent response. Price would significantly moderate the relationship between fashion involvement and quality. Opinion-seeking was found not to be a significant stimulus for Malaysian e-consumers.Originality/valueContribution of this study goes to the existing literature by providing a deeper understanding of Malaysian e-consumer behavior by applying S-O-R theory. Malaysian fashion e-consumerism was suggested to be influenced by product involvement, quality, price, opinion-seeking and product range offered; therefore, a proposed framework was demonstrated and tested.

Modeling Instant User Intent and Content-Level Transition for Sequential Fashion Recommendation

Fashion recommendation, aiming to explore specific user preference in fashion, has become an important research topic for its practical significance to the fashion business sector. However, little work has been done on an important sub-task called sequential fashion recommendation, which aims to capture additional short-term fashion interest of users by modeling the item-to-item transitions. In this paper, we propose a novel Attentional Content-level Translation-based Recommender (ACTR) framework, which simultaneously models the instant user intent of each transition and the intent-specific transition probability. Specifically, we define instant intent with the relationships between adjacent items that the users interacted, which are the three fundamental domain-specific relationships of: <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">match</i> , <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">substitute</i> and <italic xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink">others</i> . To further exploit the characteristics of fashion domain and alleviate the item transition sparsity problem, we augment the item-level transition modeling with multiple sub-transitions using various content-level attributes. An attention mechanism is further devised to effectively aggregate multiple content-level transitions. To the best of our knowledge, this is the first work that specifies the implicit user actions in online fashion shopping with explicit instant intent, which enhances the connectivity of fashion items and boosts the recommendation performance. Extensive experiments on two real-world fashion E-commerce datasets demonstrate the effectiveness of the proposed method in sequential fashion recommendation.

이커머스 후기글 평가분석 트리플에 기반한 평가주석 데이터셋 EVAD 구축 방법론

This study aims to introduce a methodology of building evaluation-annotated datasets based on the evaluation-triple schema that we extracted from the review texts of fashion e-commerce apps. The evaluation triple defined in this study consists of ‘TARGET’, ‘ASPECT’, and ‘VALUE’. We classified the pair {ASPECT-VALUE} into 35 categories, which were characterized as three categories: ‘Information-, Judgment- and Suggestion- related

Analysis Customer Satisfaction in Luxury Fashion E-commerce Platform Services Based on Kano Model

With the digitalization trend and the changing of customer preferences, various online luxury consumption platforms occurred. In the Chinese market, lots of platforms combine e-commerce and customer engagement. Among these platforms, the Dewu app was initially launched to target customers interested in trendy luxury sneakers and have high popularity in the 90's generation in China. With high potential in the online luxury fashion industry, this research will mainly focus on improving the customer satisfaction of luxury fashion e-commerce platforms. It is critical for existing or new businesses of the online luxury fashion industry to improve their services to gain a higher market share from customer satisfaction. To better analyze, this research will adopt Kano Model to analyze the Dewu app's current practices to explore how other similar businesses could improve. Additionally, this study will adopt a questionnaire to generate data. As a result, services like identification service, sale reminder, reporting of infringement, logistics services, and customer service center will be one-dimensional quality, which indicates the importance of these services for Dewu to gain a competitive advantage. On the other hand, services like maintenance, live stream, and virtual shopping are indifferent quality and unnecessary services, and an extensive product range is attractive. The finding indicates which service is highly related to and which is less related. Through the analysis of the Dewu app, this study put forward some suggestions for other luxury fashion e-commerce platforms to improve themselves.

The Effect of Social Media Content on Buying Decision of HijUp.com

Social media is a channel to socialize with each other via online so that people can interact with each other without limited space and time. A variety of business sees many opportunities in the online media that can be used as a place to sell goods and services. Social media appears one of them is Facebook, Instagram, Twitter, YouTube and other social media. Hijup, a muslim fashion e-commerce, utilizes social media such as Instagram and Facebook to perform marketing strategies for its products. This study aims to determine the effect of social media content facebook and instagram on buying decision Hijup.com, to know the direction and influence of social media content facebook Hijup who play a role in buying decision at Hijup.com. The main theory used for this research is Elaboration Likelihood Model (ELM) to explain persuasion that sees every individual becomes very rational when evaluating persuasion messages, messages received differently for each individual then the impact will also be different. This research is using quantitative methodology with positivistic paradigm and explanatory of causality approach by using survei on 100 respondents. Hypothesis test results stated that there is a positive influence between social media content facebook and social media content Instagram simultaneously to the decision of purchase Hijup. There is a positive direction of influence on facebook social media content on purchasing decisions.

Designing, Developing and Evaluating a Sustainable Shariah-Compliant E-Commerce Fashion Prototype

This paper describes the full process of design, development and evaluation of a Shariah-compliant e-commerce fashion prototype. The prototype was designed and developed based on Fogg’s (2003) Persuasive Technology principles, namely, the principle of suggestion, principle of similarity and principle of trustworthiness which are able to persuade consumers to visit and eventually purchase products from this site. User Experience Testing was conducted to evaluate user experience in terms of screen layout, features and functionality of the site. Findings of the study indicate that the information offered is useful and the website is well-designed. In addition, the suggestions on size and clothes based on users’ body measurement are helpful when choosing suitable clothes based on one’s body type. Therefore, it can be concluded that Shariah-compliant fashion on this e-commerce site is interesting, useful and user friendly, and thus, can be a sustainable go-to site of e-commerce for Muslim women.

Factors affecting continuous purchase intention of fashion products on social E-commerce: SOR model and the mediating effect

This paper proposes a study on the impact of social e-commerce fashion products on continuous purchase intention, and explores the relationship between social media interactivity, perceived value, immersion experience and continuous purchase intention. By constructing a SOR conceptual model, 850 Chinese subjects were invited to participate in the questionnaire survey, and the data of 776 valid questionnaires were analyzed. The results show that social media interactivity has a significant impact on perceived value, immersion experience and continuous purchase, perceived value has a significant positive impact on immersive experience and continuous purchase intention, and immersive experience has a significant positive impact on continuous purchase intention. Perceived value and immersive experience play a mediating role in the relationship between social media interactivity and continuous purchase intention. The gender and education differences test of the variables were statistically significant. This research has certain theoretical value and practical guidance for the marketing of social e-commerce fashion products and the design and development of social e-commerce platform.

Chatbot design approaches for fashion E-commerce: an interdisciplinary review

ABSTRACT Chatbots can bring innovation in online assistance and communication with customers. Due to the growth of e-commerce, fashion brands have been adopting chatbots to provide personalised consumer experiences. Research in the area of chatbots for fashion e-commerce has addressed technological advancements and consumer behaviour, but little has been done on analysing chatbot features through a holistic point of view. The aim of this paper is to offer an interdisciplinary review through a comprehensive categorisation of recent studies on the theme and inform future research in the area. To achieve that, a theme-based literature review was carried out through the analysis of specialised research. The collected work was categorised addressing both computational and non-computational perspectives. The findings show that Deep Learning, recommendation systems, audio recognition and integration of chatbots with other fashion applications are a few design opportunities to be applied in both research and practice.

How to Communicate Sustainability: From the Corporate Web to E-Commerce. The Case of the Fashion Industry

Sustainability strategy at companies has become a key business and management aspect for the development and success of an enterprise. The communication of strategies and actions relating to sustainability has become increasingly important for both companies and brands. This research studies the communication process that forms part of the sustainable strategy of fashion companies, ranging from the corporate website to e-commerce, and it proposes improvements for sustainability communication. Two new models are presented: the Operational Model for Evaluating Fashion Corporate Websites (OMEFCW) and the Operational Model for Evaluating Fashion E-Commerce (OMEFeC), based on the core dimensions of online sustainability communication (orientation, structure, ergonomics and content—OSEC), as established by Siano. In order to obtain an optimal view of the fashion industry, four corporate groups—two luxury fashion groups (Kering Group and Moët Hennessy Louis Vuitton (LVMH)) and two fast fashion groups (H&amp;M Group and Intidex)—are compared. In addition, all of the e-commerce operations of the groups’ fashion brands are analyzed, a total of 32 brands. The results show that it is necessary to continue improving in terms of the communication of sustainability within the fashion industry, whilst demonstrating the great deficiency that exists regarding the communication of sustainability in the case of the brands’ e-commerce operations, which are precisely the web pages most visited by consumers.

The Implementation of Sharia Marketing Communication in The Muslim Fashion E-Commerce in Indonesia

This research aimed to analyze the implementation of muamalah provisions based on Islamic law in Muslim fashion e-commerce. The case study method was utilized to achieve the objective. The theoretical approach used is marketing communication management in Islam. The research was conducted at HIJUP.COM, the first and largest Muslim fashion e-commerce in Indonesia. The data collection techniques used in this study were written documentation, digital data archive records, in-depth and focused interviews, participant observation, physical devices, and literature study. Seven informants consisted of experts and practitioners of e-commerce, sharia marketing communication, and religious leaders selected through a purposive sampling technique. The results proved that e-commerce transactions followed the general selling and purchasing contract in Islamic law. Ideally, the implementation process includes four characteristics of sharia marketing, that is Theistic (Al-Rabbaniyah), Ethical (Al-Akhlaqiyah), Realistic (Al-Waqiah), and Humanistic (Al-Insaniyah). Obstacles in implementing sharia marketing communication included the low level of understanding and kaffah and the creative processes that must be smarter because it includes two dimensions, business and the hereafter.

How virtual avatar experience interplays with self-concepts: the use of anthropometric 3D body models in the visual stimulation process

We explored how viewing one’s anthropometric virtual avatar would affect the viewer’s self-body perception through the comparative evaluation of self-concepts—self-esteem and self-compassion, within the framework of allocentric lock theory. We recruited 18 female adults, aged 18–21, who identified themselves to have some level of body image concerns, and who had had no clinical treatment for their body image. Participants were randomly assigned either to the experimental or control group. The experimental group participated in both body positivity program and virtual avatar program, whereas the control group attended the body positivity program, only. The results affirmed that the body positivity program served as a psychological buffer prior to the virtual avatar stimulus. After the virtual avatar experience, the participants demonstrated self-acceptance by lowering their expectation on how they should look like. The findings from exit interviews enriched the quantitative results. This study verified the mechanism of the altered processing of the stored bodily memory by the egocentric sensory input of virtual avatars, and offered practical potential of the study outcomes to be applied in various emerging fields where novel applications of virtual 3D technology are sought, such as fashion e-commerce.

The Da'wah Messages as the Spiritual Marketing Approach of Islamic Fashion E-Commerce at HIJUP.COM

Indonesian Muslim consumers are looking for functional and emotional benefits in products and seeking spiritual blessings. This trend is a challenge for Islamic fashion e-commerce in promoting its products. This study uses a qualitative approach with the type of single case study research methods. The subject of this research is Islamic fashion e-commerce HIJUP.com. Researchers selected informants in the study using purposive sampling and snowball sampling techniques. Researchers used two analysis methods in case of breakdowns: an analysis based on theoretical propositions and developing case descriptions. The study results show that in Islamic fashion e-commerce, HIJUP.com contains the contents of da'wah messages, precisely, Look Good, Feel Good, Do Good. The da'wah message includes a philosophy that leads to obedience to Allah SWT concept. The message structure uses conclusion drawing and order of presentation. The message display in HIJUP.com is high-tech but straightforward techniques. The message format is unique, with interests in story or narrative and visual elements. The source of the message uses brand ambassadors, influencer marketing, and also buzzers. The study concludes that the da'wah message conveyed by Islamic fashion e-commerce includes aqidah, sharia, and morals that focus on obedience to Allah SWT, not in the preaching style while inspiring to empower Muslim women.