STRG logo as displayed in the footer of the website.

MAGAZINE

Exploring the Future of E-commerce: Bridging Conceptual Modeling and Machine Learning

On an exciting journey of innovation, our very own Research and Data Science Team Lead, Dadhichi Shukla, attended the 42nd International Conference on Conceptual Modeling ER 2023 in Lisbon 🇵🇹. This prestigious event serves as a hub for thought leaders and paves the way for discussions on the latest advances and future challenges in conceptual modeling. 

KEY TAKEAWAYS

The conference delves into theories of concepts and ontologies, and techniques for transforming conceptual models into effective software implementations. Keep reading if you're interested in:

  • Insightful Conference Links about software design and ER 2023 Keynotes
  • STRG Research Breakdown with benefits for the Retail and E-Commerce industry.
  • Strategic Advantage of web-simulated environments
  • Testing Web Agents in action by booking a demo call with our team!
Conceptual Modeling & Visionary Keynotes

Conceptual modeling is the pivotal point in businesses and industries where complicated information, data and processes require careful structuring, management and optimization to achieve operational efficiency and business goals. From IT and software development to the intricacies of financial services, healthcare, manufacturing, supply chain, retail, public sector, telecommunications, consulting and media, the importance of conceptual modeling is deeply impacting and is a crucial component for a wide range of industries. 

The International Conference on Conceptual Modeling ER 2023 in Lisbon is not only a hub for computer science experts, but also a stage for exchanging relevant trends and development in the field. Let’s take a look at the exciting keynotes that are shaping the discourse in the fields of software design, conceptual modeling, language understanding and biomedical artificial intelligence.

Prof. Daniel Jackson at Massachusetts Institute of Technology (MIT), unfolds a new approach to software design. Focusing on individual concepts, he pioneers a strategy that diagnoses usability issues more effectively, stimulates innovative designs, and constructs modular apps with enhanced separation of concerns. 
Read more

Prof. Maurizio Lenzerini from Sapienza University of Rome explores the intersection of Conceptual Modeling (CM) and Knowledge Representation and Reasoning (KRR). He delves into the evolution from early data modeling to the intricacies of ontologies and Knowledge Graphs. Read more

Walid S. Saba, a Senior Principal Scientist, Experiential Institute for Artificial Intelligence, Northeastern University, explores the bottom-up reverse engineering of language. He discusses the challenges and potential of symbolic settings in achieving an explainable model of how language works. Read more

Catia Pesquita, Associate Professor at the University of Lisbon, explores the significance of negative knowledge in biomedical AI. She sheds light on the challenges faced when distinguishing between negative and unknown facts in critical domains like healthcare. Read more

Previous slide
Next slide
STRG Research Paper Explained

Our recently published research paper is presenting a solution for specific online customer experiences. With e-commerce’s meteoric rise to a multi-billion dollar industry, the demand for quick & easy online portal creation has sky-rocketed! Using modern web tools, business owners can access pre-made templates to swiftly create an online portal for their customers. The online presence becomes everything and the shopping experience is in high demand! However, reliance on pre-designed user journey templates hampers a thorough analysis of the user experience before launching your web portal of any kind.

Modern web-tools and services enable business owners to create online portals within a short period. In most cases, the business owners design their online shops based on existing user journey templates offered by webpage service providers. As a result, it limits their analysis of the user journey of their shop portal prior to the launch (go online). It becomes extremely challenging to understand what the experience of the visitors following the template user journey could be.

Our research team, led by Dadhichi Shukla has recognized this challenge and started researching a groundbreaking approach to bridge the gap between conceptual modeling and machine learning and find a way to optimize web portals ahead of launch. The paper titled “Bridging the Gap: Conceptual Modeling and Machine Learning for Web Portals” explores the intersection of conceptual modeling and machine learning to analyze the user journey of online shopping portals. In the era of e-commerce’s significant economic impact, the challenge lies in understanding user experiences before the launch of web portals, especially when based on existing web templates. The authors propose a system using conceptual modeling and reinforcement learning, employing Q-learning and SARSA methods. By training the so called web agents to navigate simulated environments representing web portals, the study establishes a bridge between conceptual modeling and machine learning through empirical analysis. The work enhances the understanding of user behaviors, contributing to the optimization of online shopping experiences. So far so good, but what does this mean? How can businesses benefit from such “simulation”?

The Benefits of Simulated Web Environments

This paper introduces a novel system, that blends conceptual modeling and reinforcement learning. Leveraging Q-learning and SARSA, our approach simulates web portals, providing thorough insights into user journey and user experience. It is tailored to transform the online shopping experience. 

Pre-Launch Analysis

Gain competitive insights into user behavior before launching your web store. The analysis of simulated environment breaks down potential user behaviors, allowing businesses to make and refine strategic decisions prior to going live.

Data-Driven Decision-Making

Dynamic as the online landscape. The proposed method facilitates data-driven decision-making. Instead of relying solely on historical or post-launch data, businesses can generate empirical data through simulated agents. This data fuels informed decisions related to user journey design, potentially resulting in more effective and user-friendly online shops.

Complex User Behaviors

E-commerce portals are often complex with interconnected pages. Simulated environments enable the training of agents to navigate and interact within this complexity, helping you understand how users might navigate through the portal and addressing potential challenges in analyzing real user data before launch.

AI Simulation

Agents, acting as AI-driven simulations of user behaviors, provide a controlled environment for predicting and understanding how users might interact with an online store. This predictive capability is instrumental in anticipating challenges and optimizing the portal's design to meet the expectations of online shoppers.

Having simulated web environments with agents is definitely a strategic advantage. It is a safe space to test, challenge, and experiment with different elements of your online shop, design blocks, and user journey pathways. It allows you to generate empirical insights based on reinforcement learning, ensuring constant process refinement and adaptability to the dynamic landscape of retail and e-commerce. Simply, it’s the new wave of retail tech.

The Research Team
Dadhichi

Dadshichi Shukla

Research and Data Science Team Lead

Eugen

Eugen Lindorfer

Registered Manager and Product Owner

69553e3d-b838-4162-9e2b-e1573afbf585

Sebastian Eresheim

Junior Researcher IT Sicherheitsforschung

3ead1776-1032-4e58-b872-2e2d321ad00b

Alexander Buchelt

Assistent Lehre und IT Forschung

The research paper is published in Advances in Conceptual Modeling, as part of the book series Lecture Notes in Computer Science (LNCS). 

This book constitutes the refereed proceedings of 7 workshops, held at the 42nd International Conference on Conceptual Modeling, ER 2023, held in Lisbon, Portugal, during November 6-9, 2023.  The 28 full and 2 short papers were carefully reviewed and selected out of 53 submissions. Topics of interest span the entire spectrum of conceptual modeling, including research and practice in areas such as theories of concepts and ontologies, techniques for transforming conceptual models into effective implementations, and methods and tools for developing and communicating conceptual models.

The conference not only showcased the evolving paradigms in software design and AI but also spotlighted the intersection of conceptual modeling across various industries, from manufacturing and logistics to finance, healthcare, and beyond. STRG’s research paper introduces strategic solutions for the retail and e-commerce industries. By blending reinforcement learning with conceptual modeling, we are introducing a radical shift in pre-launch analysis for web portals and tackling major challenges for the booming billion-dollar online sector.

GET YOUR FREE COPY

Please fill out the form to get your free copy of our most recent research paper Bridging the Gap: Conceptual Modeling and Machine Learning for Web Portals. 

STRG.agents at your service

Enhance user journey and boost conversions

Get in touch with our team to see agents in action

You may also like

Subscribe To Our Newsletter

Get notified about new articles

Thanks for showing interest in our research!