Interactive Platform for Surveys in Information Visualisation

Barbara Nascimento

1,2 a

, Jo

ao M. Cunha

2 b

and Evgheni Polisciuc

2 c

Institute for Interdisciplinary Research, University of Coimbra, Portugal

CISUC/LASI – Centre for Informatics and Systems of the University of Coimbra, Department of Informatics Engineering,

University of Coimbra, Portugal

{barbararamos, jmacunha, evgheni}@dei.uc.pt

Keywords:

Literature Browser, Web-Based Systems, Interactive Surveys, Survey Analysis, Information Visualisation.

Abstract:

Conventional literature review platforms typically present data in static formats, restricting users from fully ex-

ploring complex, multivariate information and hindering their ability to stay up-to-date with evolving research.

A way to address this involves the development of interactive platforms that enable users to dynamically ex-

plore data and the addition of new information. In this paper, we analyse existing platforms, identifying key

features that improve data visualisation and interaction. Our analysis is used to support the proposal of a novel

interactive survey platform aimed at improving the analysis and exploration of bibliographic data using various

visualisation and interactivity techniques. The proposed platform integrates data representation approaches,

such as grids, distribution charts, and tree diagram, and interaction techniques like ﬁltering, selecting, and re-

conﬁguring data layouts. To assess the usability of the platform, we applied it to a speciﬁc scenario – a survey

on data glyphs – and conducted a study with users. The results of the user study indicate that the platform’s

functionalities, which include various representation approaches and interactive techniques, enable deep ex-

ploration and effectively assist in conducting data analysis tasks. The platform is a promising step towards

creating more dynamic and interactive survey tools and it serves as a starting point for further development.

1 INTRODUCTION

Surveys and literature reviews are essential to the ad-

vancement of academic research, especially in dy-

namic and quickly-evolving areas such as information

visualisation, where new techniques and applications

are constantly emerging. Several examples include

the use of machine learning for visualisation (e.g.,

generation and evaluation of visual models), the use

of visualisation of explainability and trustoworthines

of AI models, among many others.

Such surveys most often consist of constructing

and analysing multidimensional dataset of diverse

types of attributes that characterise the area of fo-

cus. One of the main challenges is the complexity

of analysing and visualising the multivariate data that

surveys often involve. A promising solution to over-

come these limitations is the development of dynamic

interactive survey platforms. These platforms facil-

itate navigation through the data, enabling the user

to explore different levels of detail according to their

https://orcid.org/0000-0002-4185-0074

https://orcid.org/0000-0001-6502-3500

https://orcid.org/0000-0001-9044-2707

needs. In addition, they allow survey data to be up-

dated, making it possible to maintain its relevance and

ensure that it remains up to date for longer.

In recent years, several interactive survey plat-

forms have been proposed, addressing the topics of

analysis such as biological data visualisation (Kerren

et al., 2017), hierarchies (Schulz, 2011), text visu-

alisation (Kucher and Kerren, 2015), ﬁnancial data

(Dumas et al., 2014), 3D visualisation (Kraus et al.,

2022), or Application of Machine Learning to Visual-

isation (Wang et al., 2021).

Despite their utility, such platforms have seen lim-

ited advancement in recent years. While they pro-

vide interactive navigation among research entries,

they lack sophisticated visualisation tools necessary

for conducting in-depth analyses. The data is typi-

cally presented in basic formats, such as lists or grids,

without offering advanced visualization capabilities

that could enhance analytical insights. Consequently,

researchers are often compelled to rely on external

tools or develop custom code, further extending the

already time-intensive process of survey analysis.

In this paper, we provide a comprehensive anal-

ysis of existing interactive survey platforms, with the

718

Nascimento, B., Cunha, J. M. and Polisciuc, E.

Interactive Platform for Surveys in Information Visualisation.

DOI: 10.5220/0013190600003912

Paper published under CC license (CC BY-NC-ND 4.0)

In Proceedings of the 20th International Joint Conference on Computer Vision, Imaging and Computer Graphics Theory and Applications (VISIGRAPP 2025) - Volume 1: GRAPP, HUCAPP

and IVAPP, pages 718-729

ISBN: 978-989-758-728-3; ISSN: 2184-4321

aim to identify features that support efﬁcient literature

analyses, as well as their weaknesses and limitations.

Based on this we designed and present in this paper

our new platform. The proposed platform incorpo-

rates interaction and visualisation techniques tailored

for survey tasks. A case study on data glyph survey is

used to demonstrate the platform’s functionalities.

The main contributions of this work are:

• Analysis of Existing Platforms: We perform an

analysis of interactive survey platforms, identify-

ing their most relevant features and functionalities

and focusing on how they facilitate the visualisa-

tion and exploration of data in surveys.

• Proposal of a New Platform: Based on the in-

sights obtained, we propose a new platform inte-

grating interactivity and visualisation techniques

to analyse multivariate survey data.

• Application in a Case Study: We demonstrate the

use of the proposed platform in a case study with

data glyphs and conduct a semi-structured study

with users.

This paper is organised as follows. In section 2,

we discuss previous work on interactive survey plat-

forms in the domain of visualisation. In Section 3, we

present details of the collection and analysis of the

platforms using the PRISMA guideline. In Section 4,

we present the design requirements elicitation based

on the analysis of existing platforms. In Section 5,

we dive into the details of the design and implemen-

tation of the proposed platform. Section 6 presents a

case study on data glyphs, where we applied the pro-

posed platform to visualise the survey samples. Fur-

ther, Section 7 presents a semi-structured user study,

describing its setup and results. Finally, a summary

of the paper is made in Section 8.

2 RELATED WORK

The ﬁeld of information visualisation has grown con-

siderably in recent decades, today offering a wide

range of techniques and tools. As a consequence

of this rapid development, traditional survey articles,

which usually provide a static overview of references

and literary methods are quickly outdated (Kerren

et al., 2017). Currently, many surveys are exploring

non-static ways of presenting survey data (Jena et al.,

2020). This includes web-based platforms or, in the

case of static publications, an interactive tool is pro-

vided to assist in the presentation of the data. For

example, TimeVis (Tominski and Aigner, 2023) com-

piles time visualisation techniques, which are classi-

ﬁed into several categories, such as date, time, and

visual representation, totalling 159 entries.

Various interactive survey platforms on different

topics have been proposed. Some interactive surveys

focus on speciﬁc visualisation techniques. The tree-

vis.net (Schulz, 2011) analyses a series of visualisa-

tions in trees, considering categories such as dimen-

sionality, edge representation and node alignment.

The dataphys.org (Dragicevic and Jansen, 2012) has

a collection of more than 372 physical visualisation

entries, ordered chronologically, since the ﬁrst sam-

ple dated 5500 BC. Other examples include Dynamic

graph visualisation (Beck et al., 2014), sparklines lit-

erature (Beck and Weiskopf, 2017) and geospatial

network visualisation (Sch

ottler et al., 2021).

Interactive surveys can also focus on visualisa-

tion within a particular application domain applica-

tion domain. The BioVis Explorer (Kerren et al.,

2017) presents techniques applied to visualise bio-

logical data and displays the samples using a galaxy

metaphor. The Financevis.net (Dumas et al., 2014)

analyses more than 50 papers with ﬁnancial data vi-

sualisations. The EHR-STAR (Wang and Laramee,

2022), collects visualisations used to analyse elec-

tronic health records (EHR) and population health

records (PopHR).

When observing the existing interactive surveys,

the diversity of visualisation techniques and applica-

tion domains addressed is evident. However, one in-

teresting aspect is that the way the information is pre-

sented remains relatively uniform. In the next sec-

tions, we will perform an analysis of the form of rep-

resentation of these platforms as well as the interac-

tion techniques used.

3 SCOPE AND METHODOLOGY

This section describes the method used to collect and

analyse interactive survey platforms. Our approach

followed the PRISMA 2020 guidelines (Page et al.,

2021) to identify and categorise studies that con-

ducted systematic literature reviews (SLRs) within

the visualisation domain and presented interactive

platforms. To ensure the inclusion of relevant works,

we applied three inclusion criteria: (1) the article

should focus on the visualisation domain; (2) it should

be a systematic literature review; and (3) it should in-

clude or discuss an interactive platform that supports

the review process.

We reviewed academic databases, such as Web of

Science, ACM Digital Library and IEEE Xplore, to

ensure comprehensive coverage. The review process

involved a screening and selection phase, ensuring

Interactive Platform for Surveys in Information Visualisation

719

Figure 1: A birds-view research methodology ﬂowchart, including query strategy, screening process and snowball inclusion.

that only articles that met all three inclusion criteria

were considered for analysis. In the ﬁnal stage, we

categorised the interaction techniques and representa-

tion methods described in the selected studies. The

following subsections provide more details about the

survey strategy and categorisation structure applied in

this review.

3.1 Query Strategy

The research was formulated to ﬁnd papers conduct-

ing a systematic literature review, so keywords such

as “survey”, “state of the art”, and “systematic re-

view” were used. We also used the keyword “visu-

alisation” to always ensure the paper was within the

visualisation domain.

The following databases were considered to iden-

tify and collect relevant manuscripts: (1) IEEE

Xplore, (2) ACM Digital Library, (3) Science Direct,

and (4) Google Scholar. Additionally, we collected

data from the arXiv platform to include grey litera-

ture in our survey.

We conducted three rounds of queries in each of

the databases, always with the restriction of ﬁnding

the keywords in the title of the articles. The ﬁnal

query sequence was: (“visualisation”) AND (“liter-

ature systematic review” OR “survey” OR “state-of-

the-art”). Finally, one of the considered platforms

contained a list of other interactive surveys, which we

added to our list of surveys through snowballing re-

view.

Figure 1 presents the ﬂowchart of the presented

methodology, which provides a birds-view of the pro-

cess. In the identiﬁcation phase, we gathered 349 pa-

pers from ACM, 348 from IEEE Xplore, 51 from Sci-

ence Direct, 210 from Google Scholar, and 115 from

arXiv, totalling 1073 entries. Afterwards, we removed

107 duplicates that were present. In the screening pro-

cess, we ﬁltered the 943 articles to follow inclusion

criteria 1–3 and were left with 17 entries. We included

six papers from the snowballing and manual query, ﬁ-

nalising the screening process with 23 entries. The

results covered papers published between 2000 and

July 2024.

3.2 Categorisation

We analyse the platforms in two different aspects:

representation and interaction techniques. In the rep-

resentation category, we analyse how data is organ-

ised and viewed. This analysis aims to understand

how platforms present information to the user and

identify how these impact the analysis experience,

considering the amount of information displayed, the

need for interaction and the level of the offered visual

abstraction.

In the interaction techniques category, we em-

ployed the framework proposed by Yi et al. (Yi et al.,

2007) to categorise the survey platforms.

This framework uses seven different interaction

techniques which are as follows:

• Select: Allows users to mark items of interest and

track them during data exploration.

• Explore: Enables users to examine data subgroups

and visualise them from different perspectives.

• Reconﬁgure: Rearranges the spatial organisation

of data to facilitate comparisons and new insights.

• Encode: Lets users modify visual properties like

colour or shape to customise data representation.

• Abstract/Elaborate: Adjusts the data abstraction

level for a more general or detailed view.

• Filter: Applies ﬁltering criteria to reﬁne the

dataset and focus on relevant subsets.

• Connect: Highlights relationships and patterns

within the dataset for deeper analysis, by visu-

alise implicit or explicit connections among the

elements.

IVAPP 2025 - 16th International Conference on Information Visualization Theory and Applications

720

4 ANALYSIS OF EXISTING

PLATFORMS

The analysis follows the methodology descirbed ear-

lier in the paper. This is based on two pillars: repre-

sentation and interaction capabilities. The results are

discussed in the following section and summarised in

Table 1.

4.1 Representation

From the analysis of the platforms, we identify that

the two primary forms of representation are in list or

grid. Platforms with grid representation use thumb-

nail images to identify each sample, displaying more

information such as the title when the user moves

the cursor over the thumbnail and when the sample

is selected (e.g., the SentimentVis browser (Kucher

et al., 2018)). Most platforms with list representa-

tion already make visible more details about the sam-

ple, such as authors, year of publication and referring

categories, excluding the need to select the sample to

read the detailed content. In some of the platforms

that use list representation, the information presented

to the user is only textual, not presenting images (e.g.,

Immersive Analytics with Abstract 3D Visualisations

(Kraus et al., 2022)). The only interactive survey that

offers the option to choose the representation format

is the Physical visualisation survey (Dragicevic and

Jansen, 2012), which allows the entries to be dis-

played in a grid or list format.

Another type of frequent representation is a his-

togram, which shows the distribution of papers by

year of publication. In several approaches, such as

the one used in the performance visualisation survey

(Isaacs et al., 2014), the histogram is reactive to the

selectors, which is an important component for vi-

sual analysis. Although this kind of representation

is available in many platforms, it is usually a sec-

ondary element. Only two of the surveys analysed

showed different representation approaches: the Bio

Explorer (Kerren et al., 2017) uses spatial representa-

tion with a metaphor of galaxies, and the Image Syn-

thesis Pipeline (Tsirikoglou et al., 2020) presents a

tree visualisation of the samples.

4.2 Interaction Techniques

In survey platforms, an important component is the

interaction that enables the exploration and analysis

of the literary collection. To assess interaction we re-

lied on the interaction taxonomy by (Yi et al., 2007)

described earlier.

• Filter is the only interaction technique present in

all the platforms analysed, which allows the user

to change the group of data presented based on

certain criteria, such as the category being anal-

ysed.

• Abstract/Elaborate is one of the most present

interaction techniques among the platforms anal-

ysed. The abstraction/elaboration consists of pre-

senting a popup card with details of the selected

sample.

• Select allows the user to mark items of interest

and accompany them in the exploration. The only

platforms that provide the selection interaction

technique are the BioVis Explorer (Kerren et al.,

2017) and the platforms that use the SurVis sys-

tem (Beck et al., 2014; Isaacs et al., 2014; Liu

et al., 2017; Beck and Weiskopf, 2017; Sperrle

et al., 2021). SurVis (Beck et al., 2016) is a system

for literary collection that presents a word-size vi-

sualisation of the keywords used in the literary

collection. The system has selectors that, when

applied, highlight all the keywords that were used

in conjunction with that selector, along with a

sparkline graph showing the number of times each

keyword was used. The BioVis Explorer (Kerren

et al., 2017) highlights the selected sample and a

connection to all the most similar samples when

the user interacts with a sample by pressing the

right mouse button.

• Connect highlights associations and relationships

between the data presented. The only platform

that presented this interaction technique is the

BioVis explorer (Kerren et al., 2017), which ad-

ditionally employs spatial representation from a

metaphor of galaxies based on the dimensionality

reduction.

• Explore is the interaction technique that allows

the user to examine different subgroups of data

and visualise the data from a different perspec-

tive. The only survey that presents this type

of technique is the Bio Explorer (Kerren et al.,

2017) which uses zooming and panning to navi-

gate through the visualisation.

• Reconﬁgure interaction technique was found in

two platforms (Wang and Laramee, 2022; Kraus

et al., 2022) for enabling the user to sort the sam-

ples.

• Encode was found in none of the analysed sur-

veys analysed.

Based on our analysis, we have identiﬁed several

gaps in the current platforms that can be explored.

Some platforms offer visualisations of bibliographic

Interactive Platform for Surveys in Information Visualisation

721

Table 1: Assessment of related systems in relation to the form of representation used and the presence or absence of the

interaction techniques proposed by Yi et al. (Yi et al., 2007).

Survey Representation Filter Abstract/Elaborate Select Connect Explore Reconﬁgure Encode

(Bach et al., 2017) List •

(Tsirikoglou et al., 2020) List + Tree •

(Xu et al., 2020; Sch

ottler et al., 2021) Grid •

(Beck et al., 2014; Liu et al., 2017;

Beck and Weiskopf, 2017; Sperrle

et al., 2021; Isaacs et al., 2014)

List + Bars • •

(Kucher and Kerren, 2015;

Chatzimparmpas et al., 2024; Kehrer

and Hauser, 2013; Dumas et al., 2014;

Schulz, 2011; Tominski and Aigner,

2023; Figueiras, 2013; Kucher et al.,

2018; Lu et al., 2017)

Grid • •

(Dragicevic and Jansen, 2012) Grid + List • •

(Butler et al., 2021) List • •

(Wang and Laramee, 2022; Kraus

et al., 2022)

List • • •

(Kerren et al., 2017) Scatter Plot • • • • •

material in grid and list formats, which, while use-

ful for viewing multiple entries at once, can make it

challenging to analyse the dataset as a whole. This

limitation is especially signiﬁcant in surveys, where

understanding the relationships between multiple en-

tries and their classiﬁcations is crucial.

Although certain platforms, like BioVis Ex-

plorer (Kerren et al., 2017), provide more ad-

vanced visualisations—such as galaxy-based spatial

representations—they primarily focus on visualise

items by their mutual dissimilarity using multidimen-

sional scaling (MDS) based on factors like publica-

tion year and author similarity. While this approach

is valuable in speciﬁc contexts, it does not effectively

display how variables relate or how the database en-

tries are distributed across different categories of anal-

ysis. This is particularly limiting for survey tasks,

which require a more thorough analysis, not only

based on connections but also on correlation and dis-

tribution of various database entries based on the cat-

egories being examined.

In general, the current platforms provide valuable

features for exploration (e.g., ﬁltering, abstracting,

representation in grids). However, we believe that a

platform that uses advanced forms of representation

and offers a broader range of interaction techniques

is essential for a more complete and efﬁcient user ex-

perience. Further, using appropriate tools for survey

query can streamline the process, allowing users to

efﬁciently apply various types of analysis. This is

the direction we have taken when developing a new

platform, which we will elaborate on in the following

sections.

4.3 Requirements

Based on the insights derived from our analysis we

have compiled a ﬁnal list of design requirements for

the development of a the platform. It is important

to mention that the requirements such as functional,

data, implementation, are out of scope of this paper.

The derived requirements focus on improving both

the representation of data and the range of interac-

tion techniques available. The following are the main

requirements for a platform proposal:

• Multiple Data Representation Options. The

proposed platform must support different forms of

data representation speciﬁc for survey tasks. The

ability to switch between representations will en-

hance ﬂexibility and cater to different user prefer-

ences.

• Interactive Filtering Mechanism. The platform

must include ﬁltering options, allowing users to

reﬁne the dataset based on speciﬁc criteria, such

as publication year, category, or keyword rele-

vance.

• Abstract/Elaborate Interaction. Users must be

able to control the level of detail displayed about

a sample, with options to expand or collapse in-

formation. This “abstract/elaborate” technique is

commonly implemented via popups or additional

IVAPP 2025 - 16th International Conference on Information Visualization Theory and Applications

722

Figure 2: Interface of the proposed platform, showing the ﬁltering panel (left) and the main panel in grid mode (right).

information cards.

• Reconﬁguration of Data Layout. Users should

have the ability to reconﬁgure the layout of vi-

sualised data, enabling them to sort or rearrange

samples based on different criteria (e.g., publica-

tion date, author, or relevance).

• Customisable Data Encoding. Although absent

in the platforms analysed, adding the ability to

change visual elements (e.g., colour, shape, or

size) based on different data attributes will en-

hance the platform’s ﬂexibility. Users should be

able to customise these aspects to ﬁt their ana-

lytical needs, making visualisations more intuitive

and insightful.

With these insights, we intend to create a plat-

form that simpliﬁes the analysis of systematic bibli-

ographic reviews, offering a clear and accessible vi-

sualisation.

5 PROPOSED PLATFORM

Having outlined the fundamental considerations for

the interactive survey platform, we will now describe

the proposed platform, focusing on its design and

functionalities. This section will provide a detailed

overview of how the platform is structured.

The platform consists of two panels: a left-side

panel with ﬁltering options, and the main panel,

where the database entries are displayed (see Fig. 2).

The initial representation mode shows entries as

thumbnails of representative images, each with part

of its title below, arranged on a grid. In addition to the

grid approach, the user can explore the entry collec-

tion using two other representation approaches: distri-

bution charts (Fig. 3, (b-d)) and tree diagram (Fig. 3,

(e)).

The second representation option is based on dis-

tribution charts, where a different chart is selected ac-

cording to the selected variables as seen in Figure 3.

The user can choose the attributes for the x-axis, the

y-axis and the colour through three dialogue boxes in

the upper right corner of the screen. When selecting

only the x-axis, a bar chart appears; when selecting

the x-axis and y-axis, a bubble chart appears; when

the x-axis and colour are selected, a stacked bar chart

is presented; and when attributes for the x-axis, y-

axis and colour are selected, a scatterplot is presented.

This approach aims to facilitate data analysis and vi-

sualises the distributions between the different cate-

gories analysed.

The third representation option is a tree diagram

(see Fig. 3, (e)). This visualisation provides the user

with a comprehensive view of all publications and

clearly highlights the distribution of entries between

the different categories.

Regarding the interaction techniques, we imple-

mented the following:

• Filter: the ﬁltering of the data is performed

through the side panel (e.g. (Fig. 2), where there

is a word cloud of the tags with the different cat-

egories to be ﬁltered. The tags also indicate the

frequency with the number of entries classiﬁed in

the tag reference. In addition, the side panel has

a histogram showing the frequency distribution of

data entries divided by the year of publication.

• Abstract/Elaborate: the user can see more or less

details of each publication by selecting it. When a

publication is selected, a popup with a card is pre-

sented, showing information such as the full title,

year of publication and all the categories analysed.

• Select: the platform allows the selection of publi-

Interactive Platform for Surveys in Information Visualisation

723

Figure 3: Platform interface displaying different visualisations. (a) Full-screen view of the interface with the bar chart

visualisation. Below, close-up views highlight the other available visualisation options: (b) stacked bar chart, (c) bubble chart,

(d) scatter plot, and (e) tree diagram.

cations. When a publication is selected by press-

ing the right mouse button, it is highlighted, facil-

itating identiﬁcation during the exploration of the

data, as seen in Figure 4, either through ﬁltering

or reconﬁguration.

• Connect: associations and relationships between

data entries can be observed by using the tree di-

agram approach (Fig. 3), where the distribution is

made through the different nodes and links.

• Explore: this interaction technique is available

through the possibility of zooming and panning

in the graph view (Fig. 3).

• Reconﬁgure: this interaction technique is em-

ployed as the option to sort the samples in the

ﬁrst representation panel, where the publications

are arranged in a grid with thumbnails. In addi-

tion, the user can reconﬁgure the second panel by

changing the layout of the chart by choosing the

attributes of the x-axis, y-axis and the colour, see

in Figure 3.

The platform’s design and functionality have been

carefully thought out to meet the considerations iden-

tiﬁed in the previous section. With a variety of repre-

sentation and interaction options, the platform aims to

provide the user with an efﬁcient and comprehensive

way of exploring the data, wich will be demonstrated

in the next section.

6 CASE STUDY: DATA GLYPHS

To explore the functionality of the proposed platform,

we conducted a case study on data glyph design. Data

glyphs are composite graphic objects that use visual

and geometric attributes to encode multidimensional

data (Borgo et al., 2013). There are several types of

data glyphs with different designs and concepts, from

pictorial to abstract representations, such as the Cher-

noff face and its variations (Chernoff, 1973; Ramos

et al., 2023) or the Star Glyph (J.H. et al., 1972).

The objective of the survey is to collect a series of

application studies in data glyphs, analyse the differ-

ent visual characteristics of the data glyphs and their

application context, and compare the different design

proposals. It is worth mentioning that the case study

presented in this paper is part of an ongoing research

project on data glyphs (Ramos et al., 2023; Cunha

et al., 2018) and it is only brieﬂy described for illus-

trative purposes. A more detailed report will be made

available in the near future.

6.1 Data Glyphs Classiﬁcation

In the analysis of data glyphs, we classiﬁed the glyphs

based on their visual characteristics. First, we looked

at their dimensionality, classifying them as either 2D

or 3D. Then, we examined their visual approach,

IVAPP 2025 - 16th International Conference on Information Visualization Theory and Applications

724

Figure 4: Sample selection and highlight in stacked bar

chart visualisation, showing the approach (abstract, literal

and metaphoric) on the x-axis and dimensionality with

colour (2D as green and 3D as purple).

identifying whether they were abstract or pictorial

and, if pictorial, whether they used literal, metaphor-

ical, or metonymy ﬁgures. In addition, we classiﬁed

glyphs by type, such as faces, stars, or other shapes

(Ward, 2008).

Data mapping was another aspect evaluated based

on the Ward taxonomy (Ward, 2002). We categorised

the glyphs into three data mappings: one-to-one, one-

to-many, and many-to-one. We also looked at the data

dimension, i.e., how many attributes the glyph was

encoding.

For the analysis tasks, we use Andrienko’s taxon-

omy (Andrienko and Andrienko, 2006) to classify the

glyphs into synoptic and elementary tasks. Synoptic

tasks include clustering/sorting, anomaly detection,

pattern analysis, visual search, and overview. Ele-

mentary tasks include benchmarking and data enrich-

ment.

Finally, we classify the application of glyphs in

different areas proposed by Borgo et al. (Borgo et al.,

2013), such as medical, event, multiﬁeld, geospatial,

ﬂow, and tensor visualisation. In each of these areas,

glyphs provide an intuitive and effective way to visu-

alise complex, multidimensional data.

6.2 Operations of Analysis

In research such as data glyphs, the main objective is

to understand the approaches already used and their

visual characteristics. Therefore, it is important that

the platform allows the detailed visualization of each

data entry, in this case of each glyph design collected,

and a broad view of the dataset as a whole, thus al-

lowing groupings and comparisons between samples.

We have listed ﬁve operations that both the author and

general readers may want to perform when searching

for data glyphs:

O1. Get an overview of the entire data glyph dataset.

O2. Compare different designs within a category.

O3. Group glyphs based on their similarities.

O4. Analyse possible correlations between cate-

gories.

O5. Compare glyph distributions between variables.

O1 is directly related to the analysis of the taxon-

omy used to classify the glyphs. It may be important

for the author to identify if a particular category has

a low incidence of data glyphs and can be reassessed

or even eliminated, or the opposite if there is any cat-

egory in which all glyphs are present. For example,

one of the categories analysed is the dimensionality

of the glyphs. The author may need to know if more

than 90% of the glyphs are within the 2D category or

if there is no glyph in the 3D category.

O2 is connected to O1. Authors must be able to

perceive the data set as a whole, identifying the num-

ber of entries and their distributions.

O3 deals with the possibility of a detailed view of

the glyph. As research on the design of data glyphs, it

can be necessary for the author and readers to nav-

igate between the already analysed categories and

analyse the data glyph image in the search for new

characteristics.

As in O1, in O4, it may be important to analyse

the taxonomy to evaluate possible correlations and co-

dependencies between the categories. For example,

the author may want to assess whether there is any

connection between the visual approach used in the

data glyph design and its application domain.

The case study research classiﬁes the glyphs into

eight different categories and a number of possible

variables; that is, they are multivariate data. O5 is re-

lated to the possibility of cluster formation, grouping

the glyphs based on their similarities.

6.3 Analysing Data

Within the platform, the user can perform tasks us-

ing the different visualisation panels and interactiv-

ity techniques provided. Below, we present how each

panel helps in the analysis of the survey on glyphs.

The ﬁrst panel of the platform, Figure 2, allows

the reader to get an idea of the number of entries in

the survey. With its thumbnail view, the reader can

visually identify the characteristics of the glyphs and

possible similarities with other entries when ﬁlters are

applied. For example, when we ﬁlter the glyphs and

display only those with 2D dimensionality that repre-

sent more than ﬁve attributes, they have a data dimen-

sion equal to or greater than ﬁve. Most of the glyphs

presented have a circular structure.

The distribution charts, Figure 3, allows for the

identiﬁcation of the distribution of entries within the

different categories. For example, it can be seen that

there are many more glyphs with a Many-to-One data

mapping than one-to-one. Furthermore, the possibil-

ity of reconﬁguring the graph allows the reader to ex-

Interactive Platform for Surveys in Information Visualisation

725

plore the data, thus ﬁnding possible correlations. In

the data glyph survey, we noticed that the majority

of glyphs with 3D dimensionality have a higher Data

Dimension value, being between 16 and 18.

The tree diagram representation lets the user un-

derstand the hierarchies within the categories anal-

ysed, Figure 3 (e). When viewing the tree diagram, it

is possible to see that there are many more glyphs with

abstract approach and many-to-one data mapping, ei-

ther with 2D or 3D dimensionality.

7 EVALUATION

The goal of the proposed platform is to facilitate the

analysis of survey data using visualisation. To as-

sess if the implemented interface elements and in-

teraction techniques meet this goal, we conducted a

semi-structured evaluation. The evaluation focused

on two main aspects of the platform: application of

ﬁlters and exploration of visualisations.

7.1 Setup

The semi-structured evaluation was conducted in in-

dividual sessions in a quiet and closed room. The

evaluation was conducted with 5 participants, aged

between 25 and 36 years. The group consisted of

three men and two women. The participants were

students from the computer engineering ﬁeld (2) and

from the design ﬁeld (3). Each session lasted between

25 and 45 minutes and was divided into three main

parts: (i) introduction to the platform, (ii) execution of

11 tasks by the participants, and (iii) open questions.

Participants were encouraged to think out loud during

the activities, allowing the recording of their percep-

tions and difﬁculties. The tasks were designed to sim-

ulate real scenarios of use, covering the application of

ﬁlters (T1-T5), exploration of graphs (T6-T9) and de-

tailed analysis of glyphs in the cards (T10-T11). Ex-

amples of the tasks are: “Filter the data glyphs to only

show the last ﬁve years” (T1 Filter by Year); “Create

a stacked bar chart by selecting the X variable and

colour” (T8 Stacked bar chart); and “Try to visualise

more details about a glyph. What can you discover?”

(T10 Detailing a Glyph).

At the end of each task, participants were asked

to evaluate a statement regarding the clarity and ease

of performing the task – e.g. the sentence for T10

was Finding information about the glyph was easy.

The participants were expected to answer this ques-

tion with a number on a Likert scale from 1 to 5,

where 1 would correspond to “completely disagree”

and 5 to “completely agree”. Additionally, for some

Table 2: Results of participants’ evaluations of statements

related to task ease using a Likert scale [1-5], where 1 cor-

responds to “completely disagree” and 5 to “completely

agree” – the higher the values the easier it is to complete

the task. The table presents mode (mo), median ( ˜x), average

( ¯x) and standard deviation (σ ).

Task mo ˜x ¯x σ

T1 – Filter year 5 5 4.6 0.55

T2 – Filter combination 5 5 5.0 0.00

T3 – Domain ﬁlter 5 5 4.8 0.45

T4 – Multiple Filters 5 5 5.0 0.00

T5 – Glyph type ﬁlter 5 5 4.6 0.89

T6 – Create a Bar chart 5 5 4.0 1.22

T7 – Create a Bubble chart 5 4 4.2 0.84

T8 – Create a Stacked bar chart 5 5 4.8 0.45

T9 – Analyse the Tree diagram 5 5 4.4 0.89

T10 – Detailing a Glyph 5 5 4.8 0.45

T11 – Find Oldest glyph 5 5 4.8 0.45

tasks, supplementary questions were asked to gain a

better understanding of any difﬁculties experienced.

In the third part of the evaluation, the participants

were asked open-ended questions aimed at getting an

overview of the participant’s experience and improve-

ment suggestions. The questions were: (i) How would

you describe the experience with the ﬁlters, visualisa-

tions and analysis of the cards? (ii) Was any part of

the process confusing or complicated?, and (iii) If you

could suggest an improvement, what would it be?

7.2 Results and Discussion

Table 2 summarizes the results obtained in regards to

the ease and clarity of conducting the tasks, where

higher values correspond to greater ease, presenting

mode (mo), median ( ˜x), average ( ¯x) and standard de-

viation (σ ) for all tasks. Overall, the results show that

the tasks were easy to conduct ( ¯x ≥ 5 for all tasks). In

the remainder of this section, we describe the results

obtained for each task, reporting the answers of par-

ticipants to task-speciﬁc questions and discussing the

suggestions for improvements. The section is divided

into two parts, based on the two aspects addressed in

the study – the application of ﬁlters and the explo-

ration of visualisations.

7.2.1 Filter Tasks

The ﬁrst ﬁve tasks focused on applying different ﬁl-

ters within the platform. After each task, participants

were to evaluate the ease of use and clarity of the ﬁl-

ters (Table 2).

In task T1, the participant was asked to ﬁlter the

glyphs so that only the last 5 years appear. The re-

sults indicate that participants found the ﬁlter by year

IVAPP 2025 - 16th International Conference on Information Visualization Theory and Applications

726

easy to use (mo = 5, ˜x = 5, ¯x = 4.6, σ = 0.55). Ad-

ditionally, as the year ﬁlter consists of a slider but-

ton and a histogram displaying the number of entries

for each year, participants were asked if the histogram

helped them understand the distribution of the glyphs.

Four participants answered with a score of 5 (strongly

agree), while one gave a score of 4 (partially agree),

noting that they would like the option to input the

year value directly, in addition to dragging the slider

pointer.

In T2, participants applied a ﬁlter to display only

the data glyphs with 3D dimensionality and Many-to-

One mapping. The results show that the combination

of ﬁlters was clear (mo = 5, ˜x = 5, ¯x = 5, σ = 0).

Task T3 involved ﬁltering the data glyphs by do-

main. The results demonstrated that the display of re-

sults after applying the ﬁlter was clear (mo = 5, ˜x = 5,

¯x = 4.8, σ = 0.45). Additionally, at the end of this

task, participants were asked if the results after ap-

plying the ﬁlter were clear. All participants answered

with a score of 5 (strongly agree), indicating that they

had no difﬁculty visualizing the results after applying

the ﬁlters.

In T4, which required the combination of three or

more ﬁlters, the participant could choose which ﬁl-

ter to apply. The results show that the participants

found the process of applying multiple ﬁlters simple

(mo = 5, ˜x = 5, ¯x = 5, σ = 0). At the end of the

task, participants were also asked if there any confus-

ing parts in the interface used to select multiple ﬁlters.

Three participants answered 1 (strongly disagree), in-

dicating that they had no difﬁculties, while two partic-

ipants rated 4 (partially agree) and 5 (strongly agree),

noting that the number of entries displayed in the ﬁlter

tags was confusing.

Finally, in T5, participants were asked to identify

the type of glyph with the highest number of entries.

They considered this process intuitive (mo = 5, ˜x = 5,

¯x = 4.6, σ = 0.89).

The analysis of the tasks related to the application

of ﬁlters reveals that, in general, users ﬁnd system ﬁl-

tering tools intuitive and effective when used individ-

ually. However, there is evidence that the interaction

between multiple ﬁlters can be confusing, and some

participants had difﬁculty interpreting the tags, affect-

ing their performance in Tasks T3 and T4. Addition-

ally, some participants had difﬁculty ﬁnding the total

number of data glyphs being visualised. Based on the

results of the tasks and feedback collected from par-

ticipants during open-ended questions, we have iden-

tiﬁed potential improvements to be made in future

developments of the platform. These improvements

mostly concern clarifying the interface when multiple

ﬁlters are applied simultaneously and introducing in-

structions or indicators that help guide users through

the multiple ﬁltering process. In particular, two par-

ticipants suggested the need for speciﬁc input options

within ﬁlters, such as enabling users to select or type

a year directly instead of relying solely on dragging a

slider. For tag-based ﬁlters, participants noted that the

number of entries displayed in each tag corresponded

to the total number of glyphs on the platform and

did not update dynamically when a ﬁlter was applied.

This created a misleading impression that data glyphs

existed in categories that did not match the active ﬁl-

ters. Another improvement recommended by partic-

ipants is the implementation of a ”reset” button for

ﬁlters. During the tasks, users often needed to apply

multiple ﬁlters, but the lack of an easy way to reset

them was inconvenient and negatively impacted the

user experience.

7.2.2 Visualisation Tasks

Tasks T6 to T11 were designed to evaluate the repre-

sentation approaches, speciﬁcally the various graphic

visualisations offered by the system. The goal was

to check how clear, useful, and easy is to use these

visualisations.

In T6, participants created a bar chart by select-

ing an attribute for the x-axis. The results indicated

that participants found the graph conﬁguration pro-

cess easy and intuitive (mo = 5, ˜x = 5, ¯x = 4.0,

σ = 1.22). Additionally, they were asked if the vi-

sualisation helped them understand the distribution of

glyphs; all of them answered positively that the chart

was clear and the visualisation helped them under-

stand the distribution. These ﬁndings suggest that the

bar chart interface is easy to use and provides clear

insights.

Task T7 involved creating a bubble chart by

changing the attribute on the x-axis and adding an-

other attribute to the y-axis. Participants found the

bubble graph helpful in visualising the relationship

between variables X and Y (mo = 5, ˜x = 4, ¯x = 4.2,

σ = 0.84). In addition, participants were asked if they

experienced any difﬁculty when interpreting the bub-

ble chart. One participant answered with a 1 (strongly

disagree), indicating no difﬁculties; one answered

with a 2 (partially disagree); one with a 3 (neutral);

and two with a 5 (strongly agree), indicating signif-

icant difﬁculty in interpreting the chart. All partici-

pants who reported any level of difﬁculty noted that

they struggled to identify which categories had the

highest number of entries and found it impossible to

determine the exact values.

In T8, participants created a stacked bar chart by

selecting attributes for the x-axis and colour. They

found this visualisation clear and valuable (mo = 5,

Interactive Platform for Surveys in Information Visualisation

727

˜x = 5, ¯x = 4.8, σ = 0.45). Additionally, the partici-

pants were asked if the visualisation by colours helped

analyse the distribution in the categories. All the par-

ticipants unanimously agreed, and the majority of the

participants (4) commented that this chart was easier

to analyse than the bubble chart.

Task T9 required participants to observe a tree di-

agram. The clarity of the tree diagram received posi-

tive evaluations (mo = 5, ˜x = 5, ¯x = 4.4, σ = 0.89).

In T10, participants were asked to ﬁnd informa-

tion about a speciﬁc data glyph. They reported that

it was easy to ﬁnd additional information (mo = 5,

˜x = 5, ¯x = 4.8, σ = 0.45).

Finally, Task T11 consisted of identifying the data

glyph with the oldest publication date. They consid-

ered this process intuitive (mo = 5, ˜x = 5, ¯x = 4.8,

σ = 0.45).

The analysis of the tasks related to the representa-

tion options reveals positive feedback on the usability

and clarity of the graphic visualisations offered by the

platform. Most participants found the visualisations

helpful in gaining insights into the data. However,

some areas presented challenges, such as the interpre-

tation of the bubble graph and the detailed analysis of

the tree visualisation. Additionally, participants men-

tioned a lack of captions and information about the

graphics presented. Suggested improvements include

reviewing the choice of charts, especially the bubble

chart, and improving the interface by adding a title,

a more evident caption system for the displayed at-

tributes, and indications that the chart is selectable.

8 CONCLUSION

In this paper, we present an interactive survey plat-

form that offers multiple forms of representation and

interaction techniques, aiming to provide a compre-

hensive and effective experience for the users. The

development of this platform was informed by an

analysis of existing platforms, identifying aspects to

be improved. To assess the quality of the imple-

mented platform, we used it to analyse data from a

survey on data glyphs and we conducted a user study

to evaluate the platform’s functionalities. The user

study demonstrated how the different representation

approaches, along with interactivity techniques, help

in the execution of data analysis tasks. However, for

future work, it is essential to validate the platform’s

usability with a bigger sample of users and explore

possibilities for improvement, such as the inclusion

of advanced visual coding techniques and the imple-

mentation of sophisticated data exploration methods.

With these continuous adjustments, our platform has

the potential to become a benchmark in the ﬁeld of

interactive surveys, facilitating the research and anal-

ysis of information in various areas of study. Addi-

tionally, we intend to employ this platform for other

ongoing research, such as the survey of methods for

visualisation evalutation and application of machine

learning in this domain. It is important to test the ca-

pability of the platform on a knowledge base, which

has little to none of visual content.

ACKNOWLEDGEMENTS

This work is funded by national funds through the

FCT – Foundation for Science and Technology, I.P.

(grant number 2023.02333.BD), within the scope of

the project CISUC (UID/CEC/00326/2020), and co-

funded by national funds through the FCT – Foun-

dation for Science and Technology, I.P., under the

project/support 2022.15675.CMU.

REFERENCES

Andrienko, N. and Andrienko, G. (2006). Exploratory

Analysis of Spatial and Temporal Data: A Systematic

Approach. Exploratory Analysis of Spatial and Tem-

poral Data. Springer-Verlag Berlin Heidelberg.

Bach, B., Dragicevic, P., Archambault, D., Hurter, C., and

Carpendale, S. (2017). A Descriptive Framework

for Temporal Data Visualizations Based on General-

ized Space-Time Cubes. Computer Graphics Forum,

36(6):36–61.

Beck, F., Burch, M., Diehl, S., and Weiskopf, D. (2014).

The State of the Art in Visualizing Dynamic Graphs.

EuroVis - STARs, page 21 pages.

Beck, F., Koch, S., and Weiskopf, D. (2016). Visual Analy-

sis and Dissemination of Scientiﬁc Literature Collec-

tions with SurVis. IEEE Transactions on Visualization

and Computer Graphics, 22(1):180–189.

Beck, F. and Weiskopf, D. (2017). Word-Sized Graphics for

Scientiﬁc Texts. IEEE Transactions on Visualization

and Computer Graphics, 23(6):1576–1587.

Borgo, R., Kehrer, J., Chung, D. H., Maguire, E.,

Laramee, R. S., Hauser, H., Ward, M., and Chen, M.

(2013). Glyph-based visualization: Foundations, de-

sign guidelines, techniques and applications. In Euro-

graphics (State of the Art Reports), pages 39–63.

Butler, M., Holloway, L. M., Reinders, S., Goncu, C., and

Marriott, K. (2021). Technology Developments in

Touch-Based Accessible Graphics: A Systematic Re-

view of Research 2010-2020. In Proceedings of the

2021 CHI Conference on Human Factors in Comput-

ing Systems, pages 1–15, Yokohama Japan. ACM.

Chatzimparmpas, A., Kucher, K., and Kerren, A. (2024).

Visualization for Trust in Machine Learning Revis-

IVAPP 2025 - 16th International Conference on Information Visualization Theory and Applications

728

ited: The State of the Field in 2023. IEEE Computer

Graphics and Applications, pages 1–14.

Chernoff, H. (1973). The use of faces to represent points in

k-dimensional space graphically. Journal of the Amer-

ican statistical Association, 68(342):361–368.

Cunha, J. M., Polisciuc, E., Martins, P., and Machado, P.

(2018). The many-faced plot: strategy for automatic

glyph generation. In 2018 22nd International Con-

ference Information Visualisation (IV), pages 71–77.

IEEE.

Dragicevic, P. and Jansen, Y. (2012). List of physical vi-

sualizations. www.dataphys.org/list. Last accessed

2024-04-22.

Dumas, M., McGufﬁn, M. J., and Lemieux, V. L. (2014).

Financevis. net-a visual survey of ﬁnancial data visu-

alizations. In Poster Abstracts of IEEE conference on

visualization, volume 2, page 8.

Figueiras, A. (2013). A Typology for Data Visualization on

the Web. In 2013 17th International Conference on

Information Visualisation, pages 351–358, London,

United Kingdom. IEEE.

Isaacs, K. E., Gim

enez, A., Jusuﬁ, I., Gamblin, T., Bhatele,

A., Schulz, M., Hamann, B., and Bremer, P.-T. (2014).

State of the Art of Performance Visualization. EuroVis

- STARs, page 20 pages.

Jena, A., Engelke, U., Dwyer, T., Raiamanickam, V., and

Paris, C. (2020). Uncertainty Visualisation: An Inter-

active Visual Survey. In 2020 IEEE Paciﬁc Visualiza-

tion Symposium (PaciﬁcVis), pages 201–205.

J.H., S., E.J., F., R.M., G., and H.P., F. (1972). The surgi-

cal implications of physiologic patterns in myocardial

infarction shock. pages 126–141.

Kehrer, J. and Hauser, H. (2013). Visualization and Vi-

sual Analysis of Multifaceted Scientiﬁc Data: A Sur-

vey. IEEE Transactions on Visualization and Com-

puter Graphics, 19(3):495–513.

Kerren, A., Kucher, K., Li, Y.-F., and Schreiber, F. (2017).

BioVis Explorer: A visual guide for biological data vi-

sualization techniques. PLOS ONE, 12(11):e0187341.

Kraus, M., Fuchs, J., Sommer, B., Klein, K., Engelke, U.,

Keim, D., and Schreiber, F. (2022). Immersive An-

alytics with Abstract 3D Visualizations: A Survey.

Computer Graphics Forum, 41(1):201–229.

Kucher, K. and Kerren, A. (2015). Text visualization tech-

niques: Taxonomy, visual survey, and community in-

sights. In 2015 IEEE Paciﬁc Visualization Symposium

(PaciﬁcVis), pages 117–121.

Kucher, K., Paradis, C., and Kerren, A. (2018). The State of

the Art in Sentiment Visualization. Computer Graph-

ics Forum, 37(1):71–96.

Liu, S., Maljovec, D., Wang, B., Bremer, P.-T., and Pas-

cucci, V. (2017). Visualizing High-Dimensional Data:

Advances in the Past Decade. IEEE Transactions on

Visualization and Computer Graphics, 23(3):1249–

1268.

Lu, Y., Garcia, R., Hansen, B., Gleicher, M., and Ma-

ciejewski, R. (2017). The State-of-the-Art in Pre-

dictive Visual Analytics. Computer Graphics Forum,

36(3):539–562.

Page, M. J., McKenzie, J. E., Bossuyt, P. M., Boutron,

I., Hoffmann, T. C., Mulrow, C. D., Shamseer, L.,

Tetzlaff, J. M., Akl, E. A., Brennan, S. E., Chou,

R., Glanville, J., Grimshaw, J. M., Hr

objartsson, A.,

Lalu, M. M., Li, T., Loder, E. W., Mayo-Wilson, E.,

McDonald, S., McGuinness, L. A., Stewart, L. A.,

Thomas, J., Tricco, A. C., Welch, V. A., Whiting, P.,

and Moher, D. (2021). The prisma 2020 statement:

an updated guideline for reporting systematic reviews.

BMJ, 372.

Ramos, B. N., Mac¸

as, C., Lourenc¸o, N., and Polisciuc, E.

(2023). Towards contextual glyph design: Visualizing

hearing screenings. In 2023 27th International Con-

ference Information Visualisation (IV), pages 96–102.

IEEE.

Schulz, H.-J. (2011). Treevis.net: A Tree Visualization Ref-

erence. IEEE Computer Graphics and Applications,

31(6):11–15.

Sch

ottler, S., Yang, Y., Pﬁster, H., and Bach, B. (2021). Vi-

sualizing and Interacting with Geospatial Networks:

A Survey and Design Space. Computer Graphics Fo-

rum, 40(6):5–33.

Sperrle, F., El-Assady, M., Guo, G., Borgo, R., Chau, D. H.,

Endert, A., and Keim, D. (2021). A Survey of Human-

Centered Evaluations in Human-Centered Machine

Learning. Computer Graphics Forum, 40(3):543–568.

Tominski, C. and Aigner, W. (2023). The TimeViz Browser

– A Visual Survey of Visualization Techniques for

Time-Oriented Data. https://browser.timeviz.net. Ver-

sion 2.0.

Tsirikoglou, A., Eilertsen, G., and Unger, J. (2020). A Sur-

vey of Image Synthesis Methods for Visual Machine

Learning. Computer Graphics Forum, 39(6):426–451.

Wang, Q., Chen, Z., Wang, Y., and Qu, H. (2021). A sur-

vey on ml4vis: Applying machine learning advances

to data visualization. IEEE transactions on visualiza-

tion and computer graphics, 28(12):5134–5153.

Wang, Q. and Laramee, R. (2022). Ehr star: The state-

of-the-art in interactive ehr visualization. Computer

Graphics Forum, 41(1):69–105.

Ward, M. O. (2002). A taxonomy of glyph placement strate-

gies for multidimensional data visualization. Informa-

tion Visualization, 1(3-4):194–210.

Ward, M. O. (2008). Multivariate data glyphs: Principles

and practice. In Handbook of data visualization, pages

179–198. Springer.

Xu, K., Ottley, A., Walchshofer, C., Streit, M., Chang, R.,

and Wenskovitch, J. (2020). Survey on the Analysis of

User Interactions and Visualization Provenance. Com-

puter Graphics Forum, 39(3):757–783.

Yi, J. S., Kang, Y., Stasko, J., and Jacko, J. (2007). Toward

a Deeper Understanding of the Role of Interaction in

Information Visualization. IEEE Transactions on Vi-

sualization and Computer Graphics, 13:1224–1231.

Interactive Platform for Surveys in Information Visualisation

729