Unraveling the Impact of Visual Complexity on Search as Learning
Wolfgang Gritz, Anett Hoppe, Ralph Ewerth · 2025
Read Unraveling the Impact of Visual Complexity on Search as Learning by Wolfgang Gritz, Anett Hoppe, Ralph Ewerth (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2501-05289.txt.
Introduction
Gritz, Hoppe, and Ewerth ask in twenty twenty-five how the visual clutter of web pages affects people who search in order to learn. Searchers judge pages within a second of seeing them, past work shows. Yet little was known about how clutter shapes learning itself, a gap that matters for building search tools that truly support study.
They modelled page clutter and beauty with a wide set of layout traits. They linked those traits to behaviour in a lab dataset where people learned how thunderstorms form. Fit with the topic was the strongest driver of learning gains. Still, sessions with simpler pages led to better learning, and layout traits mattered more than simple counts like image numbers. They share their code openly.
Catalog record
| CREATOR | Wolfgang Gritz, Anett Hoppe, Ralph Ewerth |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 35 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2501-05289
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Unraveling the Impact of Visual Complexity on Search as Learning Wolfgang Gritz, Anett Hoppe, Ralph Ewerth Abstract Information search has become essential for learning and knowledge acquisition, offering broad access to information and learning resources. The visual complexity of web pages is known to influence search behavior, with previous work suggesting that searchers make evaluative judgments within the first second on a page. However, there is a significant gap in our understanding of how visual complexity impacts searches specifically conducted with a learning intent. This gap is particularly relevant for the development of optimized information retrieval (IR) systems that effectively support educational objectives. To address this research need, we model visual complexity and aesthetics via a diverse set of features, investigating their relationship with search behavior during learning-oriented web sessions. Our study utilizes a publicly available dataset from a lab study where participants learned about thunderstorm formation. Our findings reveal that while content relevance is the most significant predictor for knowledge gain, sessions with less visually complex pages are associated with higher learning success. This observation applies to features associated with the layout of web pages rather than to simpler features (e.g., number of images). The reported results shed light on the impact of visual complexity on learning-oriented searches, informing the design of more effective IR systems for educational contexts. To foster reproducibility, we release our source code 1 1 1 https://github.com/TIBHannover/sal_visual_complexity . 1 Introduction The Internet’s ubiquity and the abundance of online information have fundamentally altered knowledge acquisition. Web search engines have become indispensable tools, providing the first access point to finding relevant information and resources that enable users to increase their knowledge [ 24 , 45 ] . The process of effectively searching and judging the relevance of resources is an important part of the learning process. Therefore, it is crucial to understand the factors that influence search behavior and learning outcomes to improve the effectiveness of learning in this environment [ 32 , 37 ] . Web search sessions can encompass transactional, navigational, and informational purposes [ 7 ] . The research field Search as Learning (SAL) focuses on informational search, characterized by a user’s intent to acquire knowledge from web pages. Learning success is frequently measured as knowledge gain (KG) , defined as the change in the learner’s knowledge state after the search process [ 23 , 29 ] . The relevance of search results is intricately tied to a user’s current knowledge state and the potential KG offered by the web pages retrieved. Previous research has underscored the importance of understanding learning scopes and detecting learning needs. For example, Yu et al. [ 50 ] introduced a novel set of resource-centric features, primarily focused on the linguistic complexity of web page content to enhance the predictive models. The impact of visual complexity (VisCom) on learning outcomes, however, remains largely unexplored. According to Lazard & King [ 26 ] , we refer to visual complexity as features measuring the amount and organization of visual information in a web page. Liu et al. [ 28 ] and Reinecke et al. [ 39 ] have highlighted the impact of users’ immediate judgments about website aesthetics, particularly in relation VisCom. However, little is known about how VisCom influences the learning process. An intuitive assumption could be that web pages with high complexity might be perceived as less accessible and hinder learning, especially if the subject is unfamiliar with the topic. In this paper, we investigate the relationship between VisCom and KG during web search. We model VisCom using a diverse set of features, which can be divided into four categories: (1) HTML features : statistical measures relying on the HTML code of a web page; (2) visual features : statistical measures relying on the rendered web page; (3) layout features : measures describing the web page layout utilizing the Vision-based Page Segmentation Algorithm (VIPS) [ 1 , 8 ] ; (4) aesthetic features : measures related to the Gestalt laws adopted to web pages. Using a publicly available dataset for exploratory search [ 33 ] , we address the following research questions: (1) Can visual complexity serve as a predictor of a user’s knowledge gain? (2) Which specific visual complexity features are significant predictors of users’ knowledge gain during web-based learning sessions? The remainder of the paper is structured as follows: Section 2 sums up the literature on VisCom of web pages and the research related to SAL. In Section 3 , we detail the various methods employed to extract features from web pages and explain the process of aggregating these features for users. Section 4 describes the conducted experiments. Finally, we summarize our findings, and suggest future research directions in Section 5 . 2 Related Work Search as Learning: Search engines have typically prioritized results based on relevance, user satisfaction, and commercial aspects (e.g., advertisement placement). However, since search engines have become essential learning tools, they must also be optimized for knowledge acquisition [ 7 ] . In the field of SAL, research focuses on understanding common user behavior and improving retrieval and ranking algorithms for learning [ 19 , 40 ] , e.g., for KG prediction. Vakkari [ 44 ] surveyed features that can indicate a user’s learning needs and their potential influence on knowledge acquisition during the search process. Roy et al. [ 42 ] observed that learning occurs at different times depending on prior knowledge. This underscores the importance of modeling the knowledge state and its evolution during web browsing to understand learning processes [ 9 , 31 , 51 , 52 ] . Research on investigating features that impact KG can be divided into two main streams: The first one focuses on the learning-related user behavior, e.g., based on analysis of input queries [ 11 ] , navigation logs [ 13 ] , and other behavioral features [ 18 ] . The second research direction explores what characteristics of web resources facilitate learning [ 21 , 34 , 49 , 50 ] . For example, Syed and Collins-Thompson [ 43 ] considered document-level features to improve learning outcome for short- and long-term vocabulary learning. A psychological study found evidence that text-based web pages seem to have a more substantial influence on a user’s KG [ 35 ] . In this context, Ghafourian et al. [ 20 ] and Gritz et al. [ 21 ] examined readability metrics and textual complexity regarding their connection to user behavior and their influence on KG prediction. For their prediction task, Yu et al. [ 50 ] utilized a comprehensive set of features, which included HTML statistics. Finally, Otto et al. [ 34 ] studied the effect on KG prediction, when readability and linguistic features are complemented with multimedia features. These studies contribute to a comprehensive understanding of how various factors can impact and forecast learning outcomes during web search. However, the influence of VisCom on learning outcomes has been relatively little studied, presenting a significant research gap that our work aims to address. Complexity of Web Pages: The prediction of the VisCom of web pages has long been a significant research focus, particularly in the commercial sector [ 14 , 30 ] . In recent research, the primary