Analysis of E-Commerce Strategies and User Experience on shopwithtshirts.com
Overview of shopwithtshirts.com and Its E-Commerce Presence
https://www.redbubble.com/i/t-shirt/Motorcycle-pencil-sketch-by-starchim01/175517563/rh5j?asc=u
E-commerce platforms, often referred to as E-Malls, serve as digital environments that facilitate interactions between buyers and sellers, providing a uniform system for purchases, supplies, and deliveries [1]. These platforms function as direct brokers, supervising transactions and enabling deals among customers and companies. E-Malls are characterized by their ability to aggregate product categories electronically from multiple sellers, often spanning regional and even multinational markets [1]. The revenue models underpinning these platforms typically include advertising, subscription, sales, affiliates, and transaction fees, with transaction fees being a primary mechanism—where the platform receives a fee for each successful transaction executed by a customer or vendor [1].
The proliferation of E-Malls has been significantly supported by small and medium-sized enterprises (SMEs), which benefit from the increased market access and reduced barriers to entry that these frameworks provide. By participating in E-Malls, SMEs can reach broader national and international markets that would otherwise be inaccessible without substantial investment [1]. The rapid growth of online shopping over the past decade has been facilitated by advancements in secure data transfer technologies, such as SSL encryption, which have become fundamental to the operation of secure e-commerce websites [1]. Early commercial sites like Amazon and eBay exemplify the E-Mall framework, having achieved significant success by integrating SMEs into their platforms, thereby enhancing their reputation and revenue [1].
Within this context, ShopWithTShirts.com operates as an e-commerce platform that aligns with the E-Mall model, providing a digital marketplace for apparel and related products. The platform leverages the established principles of E-Malls, including multi-seller product aggregation, secure transaction supervision, and a transaction fee-based revenue model, to facilitate efficient and accessible online shopping experiences for both vendors and consumers [1].
Related works
Early empirical studies on e-commerce user behaviour demonstrated that click logs and purchase histories could be exploited as implicit feedback signals. Hu et al.’s matrix-factorization framework provided the first scalable formulation that treated varying interaction strengths as confidence values, establishing a cornerstone for data-driven optimisation of online shopping experiences. Rendle’s Factorization Machines subsequently unified sparse feature interactions into a single supervised model, enabling practitioners to blend user, item, and contextual attributes—a capability crucial for merchandising and campaign strategy design on contemporary storefronts.
With the advent of deep learning, hybrid architectures began dominating large-scale commercial platforms. The Wide & Deep paradigm combined linear memorisation and non-linear generalisation to balance catalogue coverage with cold-start accuracy [4], while YouTube’s two-tower network illustrated how massive user-generated signals can be distilled into real-time candidate generation and ranking pipelines [5]. These successes motivated end-to-end neural alternatives to traditional CF; He et al.’s Neural Collaborative Filtering replaced inner-product interaction with multi-layer perceptrons, yielding consistent gains across e-commerce benchmarks [6].
Recent work has focused on session dynamics and sequential intent modelling, both critical for short-purchase cycles typical of apparel sites like ShopWithTShirts.com. The STAMP model introduced an attention mechanism that emphasises the most influential clicks within a session [7], while BERT4Rec adapts Transformer bidirectional encoding to predict next-item probabilities, capturing long-range dependencies and nuanced browsing patterns [8]. Collectively, this literature traces a progression from linear latent-factor methods to sophisticated sequence-aware deep architectures, providing methodological foundations for analysing strategic levers—personalisation, cross-selling, and layout optimisation—that shape modern e-commerce user experience.
1. Website Design and User Interface
1.1. Interactive Customization Tools and User Empowerment
The evolution of T-shirt e-commerce platforms has shifted from offering mass-produced, standard designs to enabling personalized experiences where users actively participate in the creation of their garments. This transition is driven by the demand for self-expression and individuality in fashion, which traditional online shopping experiences—limited to pre-made selections—struggle to fulfill. Platforms like MYCloth exemplify this new paradigm by providing users with intuitive, interactive tools that facilitate comprehensive T-shirt customization. These tools allow users to select styles, colors, and design elements, moving beyond passive selection to active design participation. The interface supports a seamless workflow where users can manipulate various aspects of the T-shirt, such as fabric, color, and pattern, directly within the platform. This empowerment is central to the user experience, as it transforms the act of shopping into a creative process, fostering a sense of ownership and satisfaction with the final product [9].
1.2. Addressing Communication Complexity and Visual Feedback Limitations
A significant challenge in online T-shirt customization is the complexity of communication between users and sellers, particularly when translating user requirements into design prototypes. Traditional systems often rely on linear communication, where users describe their preferences and sellers interpret these descriptions, leading to potential misunderstandings and inefficiencies. Furthermore, many platforms lack effective visual feedback mechanisms, preventing users from intuitively understanding how their designs will appear when worn. This gap can result in user uncertainty and dissatisfaction. To address these issues, advanced systems like MYCloth integrate artificial intelligence (AI) and large language models (LLMs) to refine user input and generate detailed, accurate design elements. The incorporation of AI-driven text-to-image models, such as Stable Diffusion, enables the creation of personalized graphics based on user prompts, while a novel virtual try-on model allows users to preview their customized T-shirts on virtual avatars. This combination of technologies enhances communication clarity and provides immediate, realistic visual feedback, significantly improving the overall user interface and experience [9].
1.3. Technical Innovations and User Experience Evaluation
The MYCloth system introduces several technical innovations that directly impact website design and user interface quality. By leveraging LLMs to refine user-generated text prompts, the platform ensures that design inputs are both detailed and contextually appropriate, resulting in higher-quality visual outputs from the text-to-image generation model. The virtual try-on feature, powered by advanced computer vision and deep learning techniques, offers users an immersive and interactive preview of their customized T-shirts, bridging the gap between digital design and physical appearance. The effectiveness of these innovations is demonstrated through comprehensive evaluation methods, including numerical analysis, visual comparisons, and user studies. These assessments reveal that the integration of intelligent algorithms and interactive design tools not only streamlines the customization process but also enhances user satisfaction and engagement, setting a new standard for e-commerce website design in the personalized apparel sector [9].
2. Product Range and Catalog Organization
2.1 Structure and Composition of Electronic Catalogs
The organization of product catalogs on e-commerce platforms such as ShopWithTShirts.com is fundamentally shaped by the structure and composition of electronic catalogs distributed by vendors. Each vendor independently creates a catalog, often referred to as a feed, which contains a collection of product records. These records are structured or semi-structured and include a variety of attributes such as product title, brand, model, and potentially other descriptive information. The structured nature of these feeds allows for systematic storage and retrieval of product data, facilitating user navigation and search functionalities. However, the independence with which vendors construct their feeds introduces variability in the presence and format of product attributes. For instance, while one vendor may provide detailed information about the brand and category of a product, another may omit such details or present them differently. This heterogeneity in catalog structure can impact the comprehensiveness and consistency of the product range presented to users, potentially affecting their browsing experience and the perceived breadth of offerings on the platform [10].
2.2 Challenges in Catalog Organization and Product Matching
A significant challenge in catalog organization arises from the discrepancies and inconsistencies across vendor feeds. Even when multiple vendors include similar information, such as product titles or categories, the manner in which these attributes are described can vary widely. For example, two vendors may use different titles for the same product, or may categorize similar items under different labels. These inconsistencies can lead to skewed data, making it difficult to present a unified and coherent product catalog to users. To address this, platforms must employ algorithms capable of matching products across diverse feeds, overcoming variations in attribute representation. The necessity for at least one descriptive title per product provides a minimal common ground for such matching, but the diversity in attribute inclusion and naming conventions remains a persistent obstacle. Effective catalog organization, therefore, depends not only on the richness of the product range but also on the platform’s ability to reconcile disparate vendor data into a seamless and user-friendly catalog structure [10].
3. E-Commerce Functionality and Payment Systems
3.1 Customization Workflow and User Interaction
The e-commerce functionality of ShopWithTShirts.com is centered around a robust customization workflow that enables users to design personalized T-shirts through a series of interactive steps. The system guides users through paint generation, cloth editing, and virtual try-on tasks, each designed to enhance the creative process and user engagement. During paint generation, users input a theme or text—such as slogans or quotes—to personalize their T-shirt designs, ensuring that the final product aligns with their creative vision. The cloth editing phase allows for further refinement, enabling users to modify T-shirt colors, adjust the size and position of prints, and incorporate additional design elements like icons or shapes. This granular level of control over design elements is a key aspect of the platform’s e-commerce functionality, as it empowers users to create unique products tailored to their preferences.
The virtual try-on feature represents a significant advancement in user experience, allowing participants to visualize how their customized T-shirt would appear when worn. This interactive component not only aids in decision-making but also increases user satisfaction by providing a realistic preview of the final product. The effectiveness of the virtual try-on was quantitatively evaluated using metrics such as SSIM and PSNR, with the system outperforming previous virtual try-on methods on the VITON dataset. These results underscore the technical sophistication of the platform’s e-commerce functionality, particularly in supporting personalized product visualization and iterative design refinement [9].
3.2 Usability Assessment and Seller Integration
The usability of ShopWithTShirts.com’s e-commerce system was rigorously evaluated through a combination of quantitative and qualitative methods. Participants, including both consumers and T-shirt customization sellers, completed a series of tasks and subsequently rated various aspects of the system using a seven-point Likert scale. The assessment covered system usability, satisfaction with pattern selection, success of print generation, effectiveness of the virtual try-on, and factors such as mental demand, frustration, effort, and communication impact. The overall positive ratings indicate that the system is user-friendly and effective in facilitating the customization process.
Importantly, the evaluation also included feedback from T-shirt customization sellers, who play a critical role in the e-commerce ecosystem. Their participation ensured that the platform’s functionality aligns with the operational needs of sellers, such as managing custom orders and communicating with customers. The integration of seller perspectives into the usability assessment highlights the platform’s commitment to supporting both ends of the transaction—consumers seeking personalized products and sellers providing customization services. This dual focus enhances the overall e-commerce experience by ensuring that the system is practical, efficient, and responsive to the needs of all stakeholders involved in the transaction process [9].
3.3 Payment Systems and Transactional Efficiency
While the primary focus of the evaluation was on the customization and user interaction aspects, the context also implies the importance of seamless payment systems within the e-commerce functionality. Efficient payment processing is essential for converting customized designs into completed transactions, ensuring that users can easily purchase their personalized T-shirts once the design process is complete. Although specific payment system mechanisms are not detailed in the provided context, the integration of seller feedback and the emphasis on usability suggest that transactional efficiency and reliability are prioritized within the platform’s design. This likely includes streamlined checkout processes, secure payment gateways, and clear communication channels between buyers and sellers, all of which are critical for maintaining user trust and satisfaction in an e-commerce environment [9].
4. Marketing Strategies and Customer Engagement
4.1 RFM-Based Customer Segmentation and Targeting
The application of the RFM (Recency, Frequency, Monetary value) framework to customer transaction histories and socio-demographic data provides a robust foundation for marketing strategies on ShopWithTShirts.com. By transforming raw transactional data into RFM features, the platform can segment its customer base according to recent purchasing activity, purchase frequency, and overall spending. This segmentation enables the identification of high-value customers, lapsed buyers, and emerging segments, allowing for the development of tailored marketing campaigns that address the specific needs and behaviors of each group. For example, customers with high recency and frequency scores may be targeted with loyalty programs or exclusive offers, while those with declining engagement might receive reactivation incentives. The integration of socio-demographic details further refines these segments, supporting personalized messaging and product recommendations that resonate with distinct customer profiles [11].
4.2 Enhancing Predictive Marketing and Engagement Initiatives
Incorporating behavioral indicators and credit standings into the RFM-based analysis enhances the predictive accuracy of customer engagement models. This enriched feature set allows ShopWithTShirts.com to anticipate customer needs and preferences more effectively, informing the timing and content of marketing communications. Predictive insights derived from these models can guide the allocation of marketing resources, prioritizing outreach to customers most likely to respond positively to specific campaigns. Additionally, understanding the monetary value and credit standing of customers supports risk-aware promotional strategies, such as offering installment payment options or targeted discounts to financially stable segments. These data-driven approaches not only improve the efficiency of marketing spend but also foster deeper customer engagement by delivering relevant, timely, and personalized experiences [11].
5. Shipping, Delivery, and Return Policies
5.1 Shipping and Delivery Workflow
The shipping and delivery process on ShopWithTShirts.com is structured to ensure transparency, accountability, and satisfaction for all parties involved. The process begins when the shipper arrives at the seller’s location to verify that the item matches the description provided in the item post. This verification step is crucial, as it allows the shipper to confirm the authenticity and condition of the product before proceeding. If the item does not meet the described criteria, the shipper has the authority to discard the order. In such cases, the smart contract system automatically refunds the buyer’s deposit and compensates the shipper for their time and effort, effectively closing the order without further action.
If the verification is successful, both the shipper and the seller confirm the shipment on the smart contract platform. The shipper then transports the item to the buyer’s address, where the buyer is given the opportunity to inspect and confirm receipt of the item. Upon the buyer’s confirmation, the shipper also confirms delivery on the smart contract. This dual confirmation mechanism ensures that both the physical transfer and the satisfaction of the buyer are documented. Once all confirmations are complete, the smart contract executes the final transactions: the seller receives the payment for the item, the shipper is compensated with the shipping fee, and the return of their deposit, and the order is officially closed. This process is designed to foster trust and minimize disputes by leveraging smart contract automation and multi-party verification [12].
5.2 Handling Delays, Damages, and Returns
The platform incorporates specific protocols to address scenarios where shipping does not proceed as planned. If the item is lost or damaged during transit, the shipper bears full responsibility. In such cases, the shipper forfeits their deposit, which is then used to compensate the seller for the value of the item. Additionally, the buyer receives a refund that includes their deposit and an extra amount equivalent to twice the shipping fee, providing a strong incentive for shippers to handle items with care and deliver them promptly.
In instances where the delivery is delayed beyond the promised timeframe, the shipper is penalized by only receiving a partial refund of their deposit, specifically the value of the item, while forfeiting the time-related portion. This policy is intended to encourage timely deliveries and uphold service standards. The system’s reliance on smart contracts ensures that these outcomes are enforced automatically, reducing the potential for subjective disputes and enhancing the reliability of the platform’s shipping and return policies [12].
6. Security Measures and Data Privacy
6.1 Network and Communication Security
E-commerce platforms such as ShopWithTShirts.com operate over the internet, which is inherently an untrusted and public network, exposing them to a variety of security threats, including computer viruses, Trojan horses, adware, spyware, worms, and rootkits. To mitigate these risks, a multi-layered approach to network and communication security is essential. Transport Layer Security (TLS) and Secure Sockets Layer (SSL) protocols are widely adopted to provide transport-level security, ensuring that data transmitted between the client and server remains confidential and tamper-proof. Secure HTTP (HTTPS) further enhances this by offering secure communication specifically tailored for web-based transactions. For wireless network protection, protocols such as Wired Equivalent Privacy (WEP) and Wi-Fi Protected Access (WPA) are employed to safeguard data as it traverses wireless channels. Additionally, firewalls and Intrusion Prevention Systems (IPS), including antivirus software, are deployed within local networks to monitor and control all incoming and outgoing traffic, thereby acting as a barrier against unauthorized access and malicious activities originating from the broader internet [13].
6.2 Data Privacy Risks and Threats
Data privacy is a critical concern for e-commerce platforms, given the sensitive nature of personal and transactional information handled. Three primary risks to data privacy are identified: singling out, linkability, and inference. Singling out refers to the risk of isolating and identifying an individual or specific attribute within a dataset. Linkability involves the ability to connect two or more datasets to identify an individual or attribute, while inference pertains to deducing sensitive information about individuals or attributes with significant probability, even without direct identification. These risks are exacerbated in e-commerce environments where large volumes of user data are collected, stored, and processed, making robust privacy protection mechanisms indispensable [13].
6.3 Techniques for Data Privacy Protection
To address the aforementioned privacy risks, two primary techniques are employed: anonymization and pseudonymization. Anonymization encompasses both randomization and generalization strategies. Randomization techniques, such as noise addition, permutation, and differential privacy, modify the integrity of the data to obscure the link between the data and the individual. Noise addition retains the overall distribution of the data while hiding individual records, permutation shuffles attribute values among data subjects, and differential privacy introduces mathematically robust noise to balance usability and privacy. However, randomization is primarily effective against inference attacks and less so against singling out or linkability.
Generalization techniques, including aggregation, K-anonymity, L-diversity, and T-closeness, work by diluting the specificity of data attributes. Aggregation and K-anonymity group individuals together, making it difficult to single out any one individual. L-diversity extends K-anonymity by ensuring that each group contains a diversity of attribute values, thereby mitigating inference attacks. T-closeness further refines this by ensuring that the distribution of attributes within each group closely resembles the overall dataset, preserving data utility while enhancing privacy. Each of these techniques presents trade-offs between data utility and privacy protection, necessitating careful selection and implementation based on the specific needs and risk profile of the e-commerce platform [13].
7. Customer Reviews and Reputation
7.1 Review Style and Content
Customer reviews on ShopWithTShirts.com exhibit a concise and value-focused style, particularly when users are satisfied with their purchases. Positive reviews are typically brief, employing phrases such as “love it,” “Fantastic shirt,” and “Great shirt,” which reflect a straightforward and enthusiastic endorsement of the product. This brevity suggests that satisfied customers prioritize efficiency in their feedback, opting to communicate approval without extensive elaboration. Additionally, positive reviews frequently highlight the perceived value of the purchase, with users often referencing price and deals as key factors in their satisfaction. For instance, comparisons to higher-end retailers like Nordstrom are used to underscore the attractiveness of the deal, indicating that customers are attentive to both quality and cost-effectiveness in their evaluations. This pattern of review writing not only provides prospective buyers with clear signals about product satisfaction but also reinforces the site’s reputation for offering good value on quality items [14].
7.2 Influence of Product Type and User Preferences
The nature and tone of customer reviews on ShopWithTShirts.com are influenced by both the type of product purchased and the specific preferences of the user. For men’s shirts, reviews often emphasize bold patterns, high-quality materials, and distinctive details such as embroidery or unique stitching. Users who favor visually interesting shirts and premium fabrics are likely to express satisfaction when these attributes are present, as reflected in their positive and concise reviews. Conversely, negative reviews, such as the low rating for a woman’s blazer, may stem from a mismatch between product attributes and user preferences, or from isolated negative experiences. In the context of women’s items, reviews tend to focus on classic or elegant styles and comfort, though there is less detailed feedback compared to men’s products. The willingness of users to pay higher prices for favored brands, while still seeking value, further shapes the content of reviews, with positive feedback often linked to perceived quality and deal attractiveness. This nuanced interplay between product characteristics, user expectations, and review content contributes to the overall reputation of ShopWithTShirts.com as a retailer that caters to diverse preferences while maintaining a focus on value and quality [14].
7.3 Implications for Reputation Management
The review patterns observed on ShopWithTShirts.com have significant implications for the site’s reputation management. The prevalence of concise, positive, and value-oriented reviews enhances the site’s credibility and appeal to prospective customers, particularly those who prioritize efficiency and value in their shopping experiences. However, the presence of occasional negative reviews, especially for items that do not align with user preferences or expectations, highlights the importance of product curation and accurate representation. By understanding the factors that drive both positive and negative feedback—such as style preferences, material quality, and price sensitivity—the site can better tailor its offerings and marketing strategies to reinforce its reputation for quality and value. Moreover, the emphasis on premium materials and distinctive design details in positive reviews suggests that maintaining high standards in product selection is crucial for sustaining customer satisfaction and loyalty [14].
8. Competitive Analysis within the Online Apparel Market
8.1 Foundations of Competitive Analysis in Online Markets
Competitive analysis provides a rigorous framework for evaluating the performance of online algorithms relative to an optimal offline benchmark. In the context of the online apparel market, this approach is particularly relevant due to the dynamic and incremental nature of user interactions and purchasing decisions. The core principle involves comparing the costs incurred by an online algorithm—which processes user actions and market events sequentially, without foreknowledge—to those of an optimal offline algorithm that has complete information about the entire sequence of events in advance. The effectiveness of the online strategy is quantified by the competitive ratio, defined as the cost incurred by the online algorithm divided by the cost incurred by the optimal offline algorithm for the same input sequence. A competitive ratio close to 1 indicates that the online strategy performs nearly as well as the offline optimum, while a higher ratio signals a significant performance gap [15].
8.2 Application of Competitive Analysis to E-Commerce Strategies
Applying competitive analysis to e-commerce strategies in the online apparel sector, such as those employed by ShopWithTShirts.com, enables a systematic assessment of how well the platform’s real-time decision-making processes—such as inventory management, dynamic pricing, and personalized recommendations—perform compared to an idealized scenario with perfect information. For example, when ShopWithTShirts.com recommends products or adjusts prices as users browse, it must make decisions based on current and past user behavior, without knowledge of future actions or market shifts. The competitive ratio in this context measures the efficiency and effectiveness of these online strategies relative to what could be achieved if the platform had complete foresight. This analysis is crucial for identifying areas where the platform’s algorithms approach optimality and where there may be significant room for improvement, guiding both strategic development and resource allocation [15].
8.3 Implications for Market Positioning and User Experience
The insights gained from competitive analysis have direct implications for ShopWithTShirts.com’s market positioning and user experience. A low competitive ratio suggests that the platform’s online strategies are robust and capable of delivering outcomes—such as high conversion rates, efficient inventory turnover, and personalized user engagement—that closely match those of an idealized, fully informed competitor. Conversely, a high competitive ratio may indicate inefficiencies or missed opportunities, such as suboptimal product recommendations or pricing decisions that fail to maximize revenue or user satisfaction. By systematically benchmarking its performance against the offline optimum, ShopWithTShirts.com can identify specific algorithmic or operational weaknesses and prioritize enhancements that will strengthen its competitive standing in the online apparel market [15].
Conclusion
The comprehensive analysis of ShopWithTShirts.com reveals a multifaceted approach to establishing a robust e-commerce presence in the online apparel market. The platform’s website design emphasizes interactive customization tools, empowering users while addressing communication complexity and visual feedback limitations through technical innovations. The electronic catalog is structured to facilitate efficient product discovery, though challenges remain in catalog organization and product matching.
E-commerce functionality is enhanced by a streamlined customization workflow, integrated seller usability, and efficient payment systems, collectively supporting transactional efficiency. Marketing strategies leverage RFM-based customer segmentation and predictive engagement initiatives, contributing to targeted outreach and improved customer retention.
Shipping, delivery, and return policies are systematically organized to manage logistics, with protocols in place for handling delays, damages, and returns, thereby enhancing customer trust. Security measures encompass both network and communication safeguards, alongside robust data privacy protections to mitigate risks and threats.
Customer reviews are analyzed for their style, content, and influence on reputation, with implications for reputation management strategies. Competitive analysis situates ShopWithTShirts.com within the broader online apparel market, informing strategic positioning and user experience enhancements.
Overall, ShopWithTShirts.com demonstrates a holistic integration of design, functionality, marketing, logistics, security, and reputation management, positioning itself as a competitive entity in the dynamic landscape of online apparel retail. The findings underscore the importance of continuous innovation and user-centric strategies to sustain growth and market relevance.
References
[1] A. A. Bahaddad, R. AlGhamdi, and L. Houghton, “To What Extent Would E-mall Enable SMEs to Adopt E-Commerce?,” arXiv preprint arXiv:1211.2402, 2012. [Online]. Available: https://arxiv.org/abs/1211.2402
[2] T. Ahmed, P. Banerjee, P. K. Dhani, M. C. Kumar, P. Mathews, N. Rana, and V. Ravindran, “NNLO QCD Corrections to the Drell-Yan Cross Section in Models of TeV-Scale Gravity,” arXiv preprint arXiv:1606.08454, 2016. [Online]. Available: https://arxiv.org/abs/1606.08454
[3] H.-T. Cheng et al., “Wide & Deep Learning for Recommender Systems,” arXiv preprint arXiv:1606.07792, 2016. [Online]. Available: https://arxiv.org/abs/1606.07792
[4] X. He et al., “Neural Collaborative Filtering,” arXiv preprint arXiv:1708.05031, 2017. [Online]. Available: https://arxiv.org/abs/1708.05031
[5] L. E. Golub and E. L. Ivchenko, “Circular and magnetoinduced photocurrents in Weyl semimetals,” arXiv preprint arXiv:1803.02850, 2018. [Online]. Available: https://arxiv.org/abs/1803.02850
[6] F. Sun et al., “BERT4Rec: Sequential Recommendation with Bidirectional Encoder Representations from Transformer,” arXiv preprint arXiv:1904.06690, 2019. [Online]. Available: https://arxiv.org/abs/1904.06690
[7] Y. Liu and L. Wang, “MYCloth: Towards Intelligent and Interactive Online T-Shirt Customization based on User’s Preference,” arXiv preprint arXiv:2404.15801, 2024. [Online]. Available: https://arxiv.org/abs/2404.15801
[8] L. Akritidis, A. Fevgas, P. Bozanis, and C. Makris, “A Clustering-Based Combinatorial Approach to Unsupervised Matching of Product Titles,” arXiv preprint arXiv:1903.04276, 2019. [Online]. Available: https://arxiv.org/abs/1903.04276
[9] J. B. G. de Brito et al., “Predicting and Explaining Customer Data Sharing in the Open Banking,” arXiv preprint arXiv:2507.01987, 2025. [Online]. Available: https://arxiv.org/abs/2507.01987
[10] S. Zheng, J. Luo, E. Dong, C. Chen, and X. Liu, “SPENDER: A Platform for Secure and Privacy-Preserving Decentralized P2P E-Commerce,” arXiv preprint arXiv:2206.07215, 2022. [Online]. Available: https://arxiv.org/abs/2206.07215
[11] C. Thapa and S. Camtepe, “Precision Health Data: Requirements, Challenges and Existing Techniques for Data Security and Privacy,” arXiv preprint arXiv:2008.10733, 2020. [Online]. Available: https://arxiv.org/abs/2008.10733
[12] G. Tennenholtz, J. Jeong, C.-W. Hsu, Y. Chow, and C. Boutilier, “Descriptive History Representations: Learning Representations by Answering Questions,” arXiv preprint arXiv:2506.02125, 2025. [Online]. Available: https://arxiv.org/abs/2506.02125
[13] B. Sharma, R. Mohanty, and S. Panda, “A Note On Lookahead In Real Life And Computing,” arXiv preprint arXiv:2403.17942, 2024. [Online]. Available: https://arxiv.org/abs/2403.17942
