What Will I Need this for Later? Towards a Platform for the Discovery

of Intra and Inter-Module Content Relations

Lisa Anders

, Daniyal Kazempour

and Peer Kr

oger

Institute of Computer Science, Kiel University, Germany

Keywords:

Course Content Relations, Connectome, Intra- and Inter-Module Relationship.

Abstract:

The large amount of knowledge in the different academic modules of bachelor and master studies renders it

difﬁcult for students to maintain the ”big picture” view, with the consequence of having ﬁnished a module

without re-connecting to it in other attended courses, despite the existence of contextual relationships between

them. The concept proposed in this work aims to provide a remedy for this problem by providing a graph-

database-founded tool that shows all modules that also deal with that particular topic for a given keyword,

additionally revealing the relationship within each of the modules. As such it provides the means for students

and lecturers alike to discover interconnectedness among different modules, preventing an isolated view of

each module, but an overarching perspective on the content learned during the studies, fostering connections

among course content.

1 INTRODUCTION

”What will I need this for later?”, is a phrase that

is overheard and stated by the authors of this pa-

per during their time as students as well as now as

teachers. Regardless of the tone and particular situa-

tion in which this question is asked, its core semantic

remains the same: the wish to understand the rela-

tions of the currently absorbed knowledge (i.e. from

a lecture or seminar) to the knowledge of future mod-

ules or even beyond in industrial or academic set-

tings. While the primary force behind this question

is driven by curiosity, it also implicitly serves an-

other objective: to introduce structure to the masses

of content permitting the linkage of knowledge within

(intra-module relationship) and across (inter-module

relationship) different modules as illustrated in Fig.1.

To foster this approach towards the discovery of

relations within and between modules, one can be

tempted to state that this can simply be done by the

lecturers in their respective courses. While indeed in-

dividual teachers address this aspect of ”What do you

need it for later” in their lectures, this may not be en-

sured in general for the following reasons:

(a) Modules that are held by different research groups

https://orcid.org/0009-0008-2556-1601

https://orcid.org/0000-0002-2063-2756

https://orcid.org/0000-0001-5646-3299

Figure 1: Illustration of relationships within a single mod-

ule (orange arrow) and between two modules (green arrow).

do not necessarily have an excessive exchange among

each other, rendering it difﬁcult to reference related

concepts of another lecture in a different scientiﬁc re-

search ﬁeld of the same domain. As a concrete exam-

ple: addressing aspects of the cell structure of bacte-

ria within a microbiology course held by lecturers of

a research group A and establishing a link to the cell

biology course held by lecturers of research group B,

(b) The lecturers potentially follow different teaching

concepts which shift the focus towards other impor-

tant aspects,

highlight relationships within and between modules.

An additional target is to motivate students to ac-

tively search for these relationships by themselves.

Anders, L., Kazempour, D. and Kröger, P.

What Will I Need this for Later? Towards a Platform for the Discovery of Intra and Inter-Module Content Relations.

DOI: 10.5220/0012700900003693

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

In Proceedings of the 16th International Conference on Computer Supported Education (CSEDU 2024) - Volume 2, pages 573-580

ISBN: 978-989-758-697-2; ISSN: 2184-5026

573

This, however, raises the need for each student to

go through the attended modules and their sheer vol-

ume of material to discover such linkages. While

students have curiosity, we see a need to support

them in introducing structure into the large volumes

of content. This adresses a problem that has also

been previously identiﬁed in (Alsaad and Alawini,

2020). To achieve this, we propose the Related Work

Connectome Project (RWCP). It is an effort to pro-

vide a platform of related-work graphs, mindmap-like

graph structures, for each module, where relation-

ships within and between modules, denoted as con-

nectomes can be discovered by students and academic

staff alike through search queries. We further en-

hanced the scope from the connectome of modules to

the connectome of related work sections of bachelor

and master theses. The motivation behind the latter is

to enable students to discover related work and pos-

sible connections among previously written theses by

other students more efﬁciently.

The expected beneﬁts of the RWCP are for once

to simplify students’ discovery of relationships within

and between modules, augmenting and motivating

them in their studies. Furthermore, another bene-

ﬁt lies in facilitating the discovery of relationships

within modules for lecturers, in order to optimize the

structure of their lectures, permitting ﬁne-tuning of

the content of their respective modules among differ-

ent working groups. Lastly, it provides researchers

further opportunities to trace and connect content

across different working groups, or in general, re-

search domains, fostering interdisciplinary scientiﬁc

endeavors.

Throughout this work, we will use the terms

related-work graph, mindmap, concept map, or graph

synonymously, although they may be used with differ-

ent preferences among the literature of their respec-

tive domains. The same condition applies to the term

module which is used interchangeably with course or

lecture.

This position paper aims to present the idea of the

RWCP, elaborate on the related work and in this con-

text also on mindmaps, and critically discuss potential

challenges and expected beneﬁts.

2 RELATED WORK

The initial idea for the RWCP is rooted in insights

gained from student evaluations of different courses,

namely the master modules Big Data Management

and Analytics, Advanced Data Mining and Machine

Learning, Knowledge Discovery and Data Mining

and the bachelor module Database Systems. In these

modules, we introduced simple mindmaps within

the past four semesters and received thoroughly

positive feedback via the EvaSys evaluation. This

stimulated the investigation into related work to the

RWCP. At the same time, the question of which

experiences other students and educators made in

different domains and at different levels of education

with mindmaps-based approaches arose. In the

following, we elaborate on related work addressing

different aspects revolving around mindmaps and the

role they play in education.

Mindmaps in the Context of Efﬁciency and

Student Involvement. Since this position paper is

founded on the concept of mindmaps, a core question

is: Can we observe an increase in learning efﬁciency?

And do mindmaps contribute to an increased involve-

ment of students in course activities? These particular

questions have been addressed in a study by (Gagi

et al., 2019) where the authors also aim to capture

the perception of the mental effort of students in a

group with and in one without utilizing mindmaps.

Additionally, they investigate the results concerning

performance in a test at the end of the study phase for

both groups. The target age group in the conducted

study were elementary school students learning about

the subject of physics. The results show that the

group that utilized mindmaps had higher results and

at the same time exhibited a lower perceived effort

compared to the control group. A more recent work

by (Fung and Liang, 2023) also supports the idea that

collaborative mindmap construction at elementary

school level enhances student motivation and, as a

consequence, also participation.

On the Beneﬁts of Mindmaps in

Higher Education. While the previous study is

based on groups of elementary school students, in the

review paper of (Machado and Carvalho, 2020) the

authors investigate the observed effects of mindmaps

(denoted as concept maps) in the context of higher

education (undergrad students) in the literature

landscape. The authors identify that concept maps

are beneﬁcial with regards to the development of

critical thinking skills and that they facilitate trans-

ference between theoretical and practical content.

The enhancement of critical thinking skills has also

been observed independently in a study on young

children in more recent work by (Polat and Aydın,

2020) serving as a cautionary indication that the

improvement of critical thinking skills while utilizing

mindmaps can be observed independently of age.

The authors further state that the fact that concept

maps develop meaningful learning skills can be

CSEDU 2024 - 16th International Conference on Computer Supported Education

574

observed and are used by students for learning

progress and assessment as well. The aspect of

assessment in the context of higher education has

also been investigated in the domain of humanities

within the work of (Kandiko et al., 2013), where

the authors come to the same conclusion: Concept

maps can help assess the individual learning progress

of students. This insight is of special interest since

the ﬁrst feedback that we received by students after

oral or written exams, emphasized that mindmaps

prompted them to create an individual learning plan

which, in turn, acted as a means to capture the

learning process. Another insight from the authors

by reviewing papers on this topic was that concept

maps promote technology inclusion. This is indeed

an aspect that supports our concept of a platform in

the scope of the Related Work Connectome Project

(RWCP). The authors of (Machado and Carvalho,

2020) however also discovered potential challenges,

namely of integrating concept mapping in academic

practices. More speciﬁcally, among the literature it

could be observed that students exhibit difﬁculties

in concept and link selection. Alongside, students

also faced challenges with software. Regardless of

the issues across the reviewed literature, concept

maps were identiﬁed to be well accepted by students

in higher education. The challenges mentioned

motivated us to identify and discuss potential difﬁ-

culties and pitfalls of this concept as seen in Section 4.

On collaborative Multi-View Aware Concept Map

Design. Moving one step further, the authors of (Pi-

cardi et al., 2020) provide the prototype of a platform

that enables a digital and collaborative approach to

conceptualizing concept maps. Besides the design

of an online tool, a crucial aspect addressed by the

authors is the fact that different students may have

different views on a topic (in our case: module/course

content relations). This is particularly of relevance

when students create concept maps in assignments or

in the context of tutorials.

Relationship-Tailored Representation of Teaching

Content. One endeavor towards a relationship-

tailored representation of teaching content is

proposed in the concept of learning paths by (Yang

and Dong, 2017), proposing pathways through sev-

eral units of learning (UoL). Two other more recent

contributions provide automatic extraction of mind or

concept maps from videos (Vimalaksha et al., 2019)

or text slides (Atapattu et al., 2017). Contrary to the

aforementioned related work, our proposed concept

does not aim to extract mindmaps but provides the

means to query and hence to explore relations within

and between mindmaps. However the creation of

mindmaps e.g. by students is a vital exercise to

develop the skill of distilling the relevant aspects of a

module and to structure them. As a consequence, we

discuss in Section 4 the necessity to train students in

the conceptualization of mindmaps.

On the Necessity for an Inquirable Platform of

Course Content Structures. To summarize, the re-

lated work so far addresses the beneﬁts of mindmaps

and their synergetic effect when created in groups of

students. Furthermore, there is previous work that

elaborates on the aspect of relationship-tailored repre-

sentation of course content. What the related work, to

the best of our knowledge, does not provide is a plat-

form that (a) contains readily available mindmaps that

represent modules and their relationships within and

between different courses and (b) that is queryable,

meaning that it yields only parts of mindmaps and its

connections that are relevant for the user (e.g. student,

teacher, scientist). This is where we aim to propose a

concept with the RWCP.

3 THE RELATED WORK

CONNECTOME PROJECT

The very foundation of the RWCP are single graph

structures that represent a module:

Deﬁnition 1 (Related-Work Graph)

A related-work graph is a directed n-ary tree T =

(V, E) consisting of vertices V and edges E. The root

of the related work graph is denoted with Ω. The

depth d of the graph T is d = 0 at the root and in-

creases with each successor node.

Figure 2: Illustration of the decreasing content granularity

from center of the related-work graph to the leaf nodes.

In this context the semantics behind the term con-

What Will I Need this for Later? Towards a Platform for the Discovery of Intra and Inter-Module Content Relations

575

tent granularity is as follows:

Deﬁnition 2 (Content Granularity)

In a related-work graph T = (V, E), the direction

from the root node Ω (of a module) to the leaf

nodes bears the semantics of decreased granular-

ity with an increasing depth d. Here the term

granularity refers to a transition from high-level

(less detailed/coarse/granular) to low-level (more de-

tailed/ﬁne) terms of the module.

To illustrate the concept of granularity we can take

the module ”Database Systems” as the root node. Be-

ing the most high-level node (we just know that it is

about databases at this point), to its next outer node

e.g. ”SQL” being a more speciﬁc term up to ”Re-

lational Algebra” being a more detailed term in the

context of the SQL language. A schema of the granu-

larity within a related-work graph can be seen in Fig

As a concrete example, Fig 3 depicts a speciﬁc

cut-out of the module contents of ”Advanced Data

Mining and Machine Learning”, revealing the details

at lower granularity, meaning a higher detail of the

content, showcasing an exemplarily intra-module re-

lation that reveals a connection between two nodes

(items) of the mindmap that were taught in different

topics.

With these deﬁnitions as ground work, we con-

tinue to provide a deﬁnition for the concept of intra-

and inter-module relations:

Deﬁnition 3 (Intra- and Inter-Module Relation)

Given two related-work graphs T = (V

, E

) and S =

, E

). An intra-module relation r

= (v

, v

) with

, v

} ∈ V

is an edge in T that short-circuits T by

creating a cycle within T that is characterized by a

node v

having an in-degree deg(v

) > 1. An inter-

module relation r

t,s

= (v

, v

) is an edge between T

and S that is also characterized by a node v

having

an in-degree deg(v

) > 1.

Finally we can now provide a deﬁnition of a con-

nectome:

Deﬁnition 4 (Connectome)

. A connectome is a database DB = {G

, ..., G

} ∪

, ..., r

, r

1,1

, ..., r

m,n

} of at least one graph G

) up to n total graphs (n denoting the number

of graphs in a DB) of potentially different granulari-

ties and a set of intra- and inter-module relations.

With the concepts of a related-work graph, content

granularity, intra- and inter-module relation and con-

nectomes we have the means for a graph database for

students and lecturers alike to discover relationships

between modules. A ﬁrst sketch of the pipeline of the

RWCP can be seen in Fig. 4 with Neo4j as the graph

database of our choosing. In this pipeline a query is

submitted via frontend to a Neo4j database containing

the connectomes. The result of the query is returned

from the backend and visualized for the user. The

connectomes can be constructed by lecturers alone or

together with students.

4 DISCUSSION POINTS

Since this position paper represents an arguable

opinion on the idea of the Related Work Connec-

tome Project (RWCP) we deem it indispensable to

elaborate on the different challenging facets that

accompany this idea. In the following, we name and

discuss a set of questions that arise in the context of

related-work graphs and that provide a challenge in

the light of different aspects.

What are the Expected Beneﬁts of this Concept?

Structured Module Content Overview.

The RWCP establishes a systematic arrangement of

module content within dedicated graphs, facilitating a

structured and organized representation of the subject

matter.

Analytical Tool for Intra- and Inter-Module Relations.

The utilization of graph databases in RWCP serves as

an analytical tool, enabling the investigation of both

intra- and inter-module relationships. The database

structure provides the means to explore the intricacies

of connections within and between modules.

Support for Learning Processes.

The RWCP aims to assist students in their learn-

ing processes through the implementation of graph

databases and associated queries. This approach

seeks to enhance the navigability of complex con-

cepts, potentially contributing to more effective learn-

ing outcomes.

Facilitation of Module Development for Educators.

Educators can beneﬁt from RWCP as it offers a plat-

form for the ongoing development and reﬁnement

of modules. The insights derived from the graph

database can empower teachers in the iterative en-

hancement of their teaching materials.

Opportunities for Interdisciplinary Exchange.

RWCP has the potential to create an inclusive

environment for collaborative exchange among

stakeholders from diverse domains. By fostering

interactions and knowledge sharing, the proposed

approach can contribute to an enriched academic

experience.

Can this Concept not be Exercised Entirely by

Students? The platform and related-work connec-

CSEDU 2024 - 16th International Conference on Computer Supported Education

576

Figure 3: Example of a partial mindmap representing parts of the structure of the ”Advanced Data Mining and Machine

Learning” module, with annotated intra-module relation (black arrow).

tomes are provided as integral components. Students

are afforded the opportunity to engage in designated

exercises, coupled with comprehensive training,

aimed at distilling graphs from content. This training

is designed to improve and practice the identiﬁcation

of relationships within and between modules. The

overarching objective is to enhance their capacity

to discern connections across the literature/content

landscape, concurrently with the skill of compressing

lecture information into a more succinct structure.

This compression process is intended to enable more

efﬁcient learning outcomes. In a recent survey the au-

thors of (Jackson et al., 2023) concluded that among

the related literature, one insight is that the design of

mindmaps by students can reveal the understanding

of the topic and serve as a way of assessment of the

learning progress for the course instructors.

What is the Estimated Effort in Maintaining

the Platform, Especially the Data? The inquiry

pertains to the anticipated effort associated with

maintaining the proposed platform, with a speciﬁc

focus on data management considerations. The facets

under scrutiny encompass IT maintenance and the

periodic updating of software and hardware alike.

Furthermore, given the evolving nature of module

content, a crucial aspect involves regular updates

of related-work graphs. Additionally there exists a

potential opportunity, possibly internal, for teaching

staff to monitor the evolution of lecture content,

tracing the lineage of both module content and

structure. A noteworthy consequence of neglecting

the timely update of graphs is the potential erosion of

student conﬁdence in the platform’s reliability.

What About the Individual Level of

Granularity? The issue of granularity at the

individual level raises pertinent considerations within

the Related Work Connectome Project (RWCP),

warranting careful examination:

Variability in Student or Teacher Granularity.

The granularity of graphs may differ among students

and teachers, introducing potential disparities. Such

discrepancies might lead to a lack of expected similar-

ities in the graphs. However, what seems as a disad-

vantage at ﬁrst sight, is actually an opportunity: In the

work of (Nuninger et al., 2019) and (Goy et al., 2017)

the authors elaborate that different individuals may

have different views on content, and as a consequence,

in the context of mindmaps, different mindmaps may

emerge from different individuals. One objective can

be, according to (Nuninger et al., 2019), to encour-

age exchange between individuals to achieve a har-

monized view that can be thought of as the small-

est common denominator of connections. One poten-

tial outcome of this harmonization process is that in-

volved individuals mutually obtain (at least parts of)

the view of others.

Challenges in Harmonizing Granularity.

Attempts to harmonize granularity face challenges,

as quantifying granularity in a standardized manner

proves difﬁcult. This poses a hurdle in achieving a

consistent level of detail across graphs.

Annotative Solutions for Granularity Differences.

A plausible solution involves empowering both teach-

ers and students to annotate or extend graphs as they

see ﬁt. This approach accommodates variations in

granularity, making visible the differences that may

otherwise remain obscured due to individual nuances.

What Will I Need this for Later? Towards a Platform for the Discovery of Intra and Inter-Module Content Relations

577

Figure 4: Pipeline of the RWCP with the interactions between user, frontend and backend.

Potential for Graph Expansion.

An inherent risk is the potentially substantial growth

of graphs. While this can be a signiﬁcant concern

in conventional mindmaps, the impact is mitigated

within the RWCP context. Queries selectively visual-

ize relevant portions, mitigating the criticality of over-

all graph expansion.

In essence, the consideration of granularity at

the individual level within the RWCP necessitates

nuanced strategies, such as annotation and selective

visualization, to address variations without com-

promising the project’s objectives, especially in the

light that different views of individuals can impact

granularity and the established relations within and

between modules.

If We Allow Teaching Staff and Students to

Enhance or Annotate the Graphs and Their

Connections, How Can We Ensure that the Data

in the Graph Database Does not Get Broken by

e.g. Wrong Content? Addressing the potential chal-

lenge of maintaining data integrity within the graph

database when permitting teaching staff and students

to enhance or annotate graphs and connections re-

quires thoughtful consideration. Several strategies

could be employed:

Moderation by Teaching Staff.

One approach involves moderation by designated

teaching staff. However, this introduces an associ-

ated overhead, necessitating additional resources for

effective oversight.

Majority Vote by Students.

An alternative method is to implement a majority vote

system among students. Given the shared interest in

maintaining accurate and valuable content for studies,

it is presumed that the majority of students would be

motivated to uphold a certain quality standard for the

graphs.

Motivation for Quality Assurance.

Leveraging the students’ intrinsic motivation for high-

quality content, this approach not only ensures data

integrity but also fosters discussions among students.

Such interactions contribute to their preparation for

lectures and exams, furthering a collective revision of

lecture content. Additionally, it provides valuable in-

sights for teaching staff into students’ comprehension

and interpretation of the material.

Drawing Parallels with Wikipedias Editing Concept.

The concept can draw inspiration from Wikipedias

model of collaborative editing, emphasizing collec-

tive contributions to content. However, it is essential

to acknowledge and manage the potential risk of ’edit-

wars’— a scenario where conﬂicting edits may pose

challenges to maintaining data coherence.

Automated Assessment of Mindmap Quality.

This approach comes with the pre-requisite of mea-

sures that capture the quality of ”how good” a

mindmap is. A ﬁrst endeavor has been made by

(Ca

nas et al., 2015) who investigate properties by

which the quality of concept maps can be assessed.

The formalization of objective(s) that provide mea-

sures for the quality of mindmaps can then be uti-

lized in well-known classiﬁcation techniques in the

machine learning domain up to the latest state-of-the-

art deep learning frameworks.

To conclude for this aspect, the integration of

moderation, majority voting, and leveraging students’

intrinsic motivation can collectively serve to uphold

data integrity while fostering an environment of

collaborative learning and content improvement

within the Related Work Connectome Project.

Can Students Exploit the RWCP? What Would be

the Consequences and How Severe Would They

Be? Exploring the potential for students to exploit the

capabilities of the Related Work Connectome Project

(RWCP) raises critical considerations and necessi-

tates a nuanced evaluation of consequences:

Strategic Query Exploitation.

Students may be inclined to craft queries highlighting

modules with the largest overlap, strategically choos-

ing courses that minimize exposure to new content.

While initially perceived as a concern, this approach

could also reveal sequences of modules that logically

build upon one another, motivating students to pursue

a coherent academic direction, such as in the domains

CSEDU 2024 - 16th International Conference on Computer Supported Education

578

of cell biology, data mining, or quantum physics.

Identiﬁcation of Overlaps.

The RWCP may showcase substantial overlaps be-

tween modules, within the same research group or

across different groups. While potentially advanta-

geous for fostering collaboration, it prompts the need

to scrutinize whether these overlaps are essential.

This analysis can lead to productive exchanges within

or between research groups, allowing for an investiga-

tion into the necessity of the overlaps. If redundant,

this insight can guide the replacement of content with

new, potentially more pertinent teaching material.

While the potential for students to exploit the

RWCP exists, it simultaneously presents opportuni-

ties for students to strategically align their academic

pursuits and for research groups to collaboratively

assess and enhance the relevance of teaching content.

The consequences are multifaceted, necessitating a

balanced evaluation of both challenges and potential

beneﬁts.

How can Users State Queries to the RWCP

Platform? The process of formulating queries to the

Related Work Connectome Project (RWCP) platform

is a crucial aspect that warrants careful examination

within the context of user interactions:

Initial Settings-Based Querying.

Users initiate the query process by selecting speciﬁc

settings, providing a structured starting point for in-

teracting with the database. This approach allows for

a tailored exploration of relevant information based

on user-deﬁned parameters.

Investigation into Natural Language Queries.

A pertinent research question involves the feasibility

of users articulating arbitrary queries in natural lan-

guage to extract meaningful results from the RWCP

platform. This inquiry seeks to understand the extent

to which users can leverage natural language inter-

faces for querying, potentially enhancing accessibility

and usability.

Phrasing Queries for Discovery.

An essential aspect of the investigation involves un-

derstanding and teaching the formulation of queries in

the platform’s context. This effort aims to demystify

the process, motivating both students and scientists to

articulate queries effectively. By fostering an under-

standing of query phrasing, the platform encourages

users to actively explore and uncover patterns and re-

lationships within and between graphs.

The mechanism for user-driven queries in the

RWCP platform involves both structured settings-

based querying and an exploration into the potential

use of natural language. Teaching the art of query

phrasing becomes pivotal in cultivating a user base

motivated to discover intricate patterns and relation-

ships within the platform’s rich dataset.

Is a Prior Introduction to that

Platform Necessary? The necessity of a prior

introduction to the Related Work Connectome

Project (RWCP) platform is a pertinent considera-

tion, even as efforts are made to maintain simplicity

comparable to modern search engines or language

models like ChatGPT (Brown et al., 2020):

Vitality of Basic Platform Introduction.

While the overarching objective is to emulate the sim-

plicity of contemporary search engines or language

models, it remains imperative to offer a concise in-

troduction to the RWCP platform and its fundamental

usage. This introduction serves as a foundational step

to ensure that users are acquainted with the platform’s

functionalities.

Flexible Introduction Methods.

The introduction can be delivered through various

modalities, including short courses dedicated exclu-

sively to platform usage, brief overviews integrated

into all modules within a 10-15 minute timeframe, or

succinct tutorial videos. The ﬂexibility in the delivery

of this introduction caters to diverse learning pref-

erences and optimizes the accessibility of platform

guidance.

Can the Relationships Within and Between

Modules not Be Extracted Through LLMs? The

examination of whether relationships within and be-

tween modules can be effectively extracted through

Large Language Models (LLMs) (Brown et al., 2020)

is a nuanced inquiry, considering the current land-

scape of advanced language models:

Emergence of Large Language Models.

The current era has witnessed a surge in the popular-

ity of LLMs, exempliﬁed by platforms like ChatGPT.

These models have demonstrated signiﬁcant prowess

in diverse language-related tasks.

Limitations of LLMs in Capturing Relations.

However, despite their proﬁciency, LLMs may face

limitations in accurately capturing intra- and inter-

module relations. To the best of our knowledge, ex-

isting LLMs may lack the capability to generate com-

prehensive graphs depicting these intricate relation-

ships. As a small test that can be conducted by the

reader, we have asked ChatGPT if it is capable of gen-

erating mindmaps from some given lecture content, to

which it responded that it can not create visual repre-

sentation. Alternatively a set of bullet points with in-

dentations is proposed which fails to illustrate more

complex inter-module relations. Therefore, as for

now, ChatGPT can not extract or generate a mindmap

What Will I Need this for Later? Towards a Platform for the Discovery of Intra and Inter-Module Content Relations

579

from given lecture slides.

Value of Manual Discovery of Relations.

The manual discovery of relations within and be-

tween graphs, promoted by students and teachers, of-

fers valuable insights. This process enables a multi-

faceted exploration, providing diverse views on po-

tential core elements of modules. Such insights, cur-

rently beyond the reach of LLMs, contribute to a more

nuanced understanding.

5 CONCLUSIONS

In this position paper, we propose and discuss the con-

cept of the Related Work Connectome Project, an ap-

proach that is aimed to provide students and lectur-

ers alike the discovery of structures within a mod-

ule and between different modules. While on a ﬁrst

glance this concept seems like ”yet another platform”

it is founded on various well-researched aspects like

concept maps, collaborative design of concept maps,

and learning pathways. As the idea behind the RWCP

can be quickly sketched, the emerging challenges as

provided in Section 4 go far deeper and reveal inter-

esting challenges and opportunities to be discussed.

We hope that this idea paves the path for enhanc-

ing research of intra- and inter-module relations, but

ﬁrst and foremost fosters an active discussion and ex-

change on the challenges and opportunities of such an

endeavor.

ACKNOWLEDGEMENTS

The project was supported by the Fund for teaching

innovation of Kiel University. The responsibility for

the content of this publication lies with the authors.

REFERENCES

Alsaad, F. and Alawini, A. (2020). Unsupervised approach

for modeling content structures of moocs. Interna-

tional Educational Data Mining Society.

Atapattu, T., Falkner, K., and Falkner, N. (2017). A com-

prehensive text analysis of lecture slides to generate

concept maps. Computers & Education, 115:96–113.

Brown, T., Mann, B., Ryder, N., Subbiah, M., Kaplan, J. D.,

Dhariwal, P., Neelakantan, A., Shyam, P., Sastry, G.,

Askell, A., et al. (2020). Language models are few-

shot learners. Advances in neural information pro-

cessing systems, 33:1877–1901.

nas, A. J., Novak, J. D., and Reiska, P. (2015). How good

is my concept map? am i a good cmapper? Knowl-

edge Management & E-Learning, 7(1):6.

Fung, D. and Liang, T. (2023). The effectiveness of col-

laborative mind mapping in hong kong primary sci-

ence classrooms. International Journal of Science and

Mathematics Education, 21(3):899–922.

Gagi

c, Z. Z., Skuban, S. J., Radulovi

c, B. N., Stojanovi

M. M., and Gaji

c, O. (2019). The implementation of

mind maps in teaching physics: educational efﬁciency

and students’ involvement. Journal of Baltic Science

Education, 18(1):117–131.

Goy, A., Petrone, G., and Picardi, C. (2017). Personal and

shared perspectives on knowledge maps in learning

environments. In Learning and Collaboration Tech-

nologies. Technology in Education: 4th International

Conference, LCT 2017, Held as Part of HCI Interna-

tional 2017, Vancouver, BC, Canada, July 9-14, 2017,

Proceedings, Part II 4, pages 382–400. Springer.

Jackson, A., Barrella, E., and Bodnar, C. (2023). Applica-

tion of concept maps as an assessment tool in engi-

neering education: Systematic literature review. Jour-

nal of Engineering Education.

Kandiko, C., Hay, D., and Weller, S. (2013). Concept map-

ping in the humanities to facilitate reﬂection: Exter-

nalizing the relationship between public and personal

learning. Arts and Humanities in Higher Education,

12(1):70–87.

Machado, C. T. and Carvalho, A. A. (2020). Concept map-

ping: Beneﬁts and challenges in higher education. The

Journal of Continuing Higher Education, 68(1):38–

53.

Nuninger, W., Picardi, C., Goy, A., and Petrone, G. (2019).

Multi-perspective concept mapping in a digital inte-

grated learning environment: Promote active learning

through shared perspectives. In Educational Technol-

ogy and the New World of Persistent Learning, pages

114–144. IGI Global.

Picardi, C., Goy, A., Gunetti, D., Petrone, G., Roberti, M.,

Nuninger, W., et al. (2020). I learn. you learn. we

learn? an experiment in collaborative concept map-

ping. CSEDU, 2:15–25.

Polat,

O. and Aydın, E. (2020). The effect of mind mapping

on young children’s critical thinking skills. Thinking

Skills and Creativity, 38:100743.

Vimalaksha, A., Vinay, S., and Kumar, N. (2019). Hierar-

chical mind map generation from video lectures. In

2019 IEEE Tenth International Conference on Tech-

nology for Education (T4E), pages 110–113. IEEE.

Yang, F. and Dong, Z. (2017). Learning path construction in

e-learning. Lecture Notes in Educational Technology.

Springer, Heidelberg.

CSEDU 2024 - 16th International Conference on Computer Supported Education

580