Comparing Novel Recurrent Neural Networks with Artificial Neural

Networks for Predicting Mental Depression from Online Video

Gaming

P. Sampath Lakshmi and S. Kalaiarasi

Department of Computer Science and Engineering, Saveetha School of Engineering, Saveetha Institute of Medical and

Technical Sciences, Chennai, Tamil Nadu, 602105, India

Keywords: Artificial Neural Network, Disability, Mental Depression, Online Gaming Novel Recurrent Neural Network,

Internet.

Abstract: The aim of this research was to enhance the accuracy in predicting mental depression in online gamers using

the Novel Recurrent Neural Network, juxtaposed against the Artificial Neural Network algorithms. Data,

sourced from various online platforms and supplemented with recent research findings, was scrutinised with

a 95% confidence interval for mean and standard deviation, iterating the process 20 times. The Novel

Recurrent Neural Network proved more accurate with a 94% precision rate, as opposed to the Artificial Neural

Network's 91%. An Independent Sample T-test further corroborated the significant disparity in predictive

abilities between the two, highlighted by a p-value of 0.00 (p<0.05). In conclusion, within the set constraints,

the Novel Recurrent Neural Network offers a superior predictive capability for mental depression in online

gamers.

1 INTRODUCTION

The auditor's primary consideration is determining

the high level of authenticity of mental depression in

online gaming (Gackenbach and Brown 2017;

Mamun et al. 2022). The analysis of mental

depression caused by online gaming involves

conducting experiments on various examples and

with different individuals (Fazeli et al. 2020). This

study utilises MATLAB source code and multiple

samples for each participant. Predictions can be made

for bipolar disorder, persistent depressive disorder,

and major depressive disorder, among other forms of

depression (Maruta, Nazarchuk, and Denysenko

2015; Irie et al. 2022). Whilst the precise origin of

depression remains elusive, it results from a

combination of genetic and environmental factors.

Depression is typically addressed through a blend of

treatments, medications, and lifestyle adjustments.

This research draws upon 1700 articles related to

the identification of mental depression or disability,

including 350 from IEEE Xplore, 100 from

Researchgate, 900 from Google Scholar, 200 from

Hindawi, and 150 from Elsevier (Mun and Lee 2021;

AS, Vickram et al. 2013). It's vital to seek help if

experiencing depressive symptoms. Depression is a

treatable condition, and with proper care, most

individuals can alleviate their symptoms and lead

fulfilling lives (Jung, Yi, and JeongDongJin 2018). If

you suspect you might be suffering from depression,

it's essential to consult a medical professional or

mental health specialist to secure an accurate

diagnosis and treatment plan (Bonnaire and Baptista

2019). If either you or someone you know is

grappling with depression, seeking assistance from a

trusted source or mental health expert is imperative

(Dias, Barbosa, and Vianna 2018; G.R et al 2014).

Table 1: Accuracy values for novel recurrent neural

network and artificial neural network.

S NO

Novel Recurrent Neural

Network

Artificial Neural

Network

92.02

88.94

92.74

89.17

93.39

89.77

93.94

90.07

94.87

90.46

95.54

91.38

95.85

91.92

96.28

92.62

96.93

92.85

97.27

93.30

128

Lakshmi, P. and Kalaiarasi, S.

Comparing Novel Recurrent Neural Networks with Artiﬁcial Neural Networks for Predicting Mental Depression from Online Video Gaming.

DOI: 10.5220/0012598700003739

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

In Proceedings of the 1st International Conference on Artiﬁcial Intelligence for Internet of Things: Accelerating Innovation in Industry and Consumer Electronics (AI4IoT 2023), pages 128-132

ISBN: 978-989-758-661-3

The survey identified a gap in research. Although

many methods for detecting mental depression or

disability exist, most have limited success rates. The

aim of this study is to enhance the accuracy of mental

depression detection by employing both Novel

Recurrent Neural Networks and Artificial Neural

Networks algorithms.

2 MATERIALS AND METHODS

NOVEL RECURRENT NEURAL NETWORKS

A Novel Recurrent Neural Network (RNN) operates

on the principle of retaining memory from previous

stages. Imagine trying to predict the next word in a

sequence without knowing the preceding context; it

would be quite tricky. Accurate predictions hinge on

this context. Traditional neural networks often

operate under the assumption that inputs and outputs

are entirely separate entities. However, in many real-

world scenarios, this isn't the case.

To address this, RNNs were introduced. Central

to the RNN's design is the Hidden Layer, which plays

a pivotal role in overcoming this challenge. The

hidden state, intrinsic to the Novel RNN, stands out

as its most defining characteristic, pivotal for

retaining sequential information.

Procedure

Step 1: Load a dataset comprising a collection of job

adverts.

Step 2: Carry out pre-processing on the dataset to

cleanse and ready the data for analysis.

Step 3: Extract a set of features from the pre-

processed dataset.

Step 4: Choose a classification algorithm and use it to

categorise the job adverts based on the extracted

features.

Step 5: Determine the accuracy of the classification

algorithm by assessing its performance on ten

randomly selected samples from the dataset.

2.1 Artificial Neural Networks

Artificial neural networks have garnered interest as a

potential diagnostic and therapeutic tool for

predicting mental illnesses, including depression.

Given their capacity to manage and analyse vast

amounts of data, such as patient details and various

clinical data, Artificial Neural Networks are aptly

suited for this purpose. One application of Artificial

Neural Networks in the realm of mental health is

predicting a patient's likelihood of developing

depression. For instance, a study utilising Artificial

Neural Networks to assess data from a large patient

cohort, as published by the Journal of Affective

Disorders, found that the algorithm was remarkably

precise in predicting the onset of depression or

disability.

2.1.1 Procedure

Step 1: Define the input and output datasets for the

neural network.

Step 2: Pre-process the data.

Step 3: Initialise the weights and bias for the artificial

neural network.

Step 4: Iterate over the training data.

Step 5: Test the Artificial Neural Network on the

dataset to assess its performance.

Step 6: Generate accuracy values for ten samples.

Table 2: Group statistics displaying the mean and standard

deviation were 94% and 1.79858, 91% and 1.58348

respectively.

GROUP NAME

Mean

Standard

Deviation

Standard Error

Mean

Accuracy

Novel Recurrent

Neural Networks

94.88.

1.79

.56

Artificial Neural

Networks

91.04

1.58

.50

2.1.2 Statistical Analysis

The study examined independent variables associated

with online gaming disorders. The dependent

variables in this context are accuracy, game hours,

and platform. Among the independent variables are

the search for excitement and basic psychological

needs. The results of the novel recurrent neural

network were compared with those of the artificial

neural network using the t-test (Mun and Lee 2021).

3 RESULTS

Table 1 provides a comparison of accuracy between

an artificial neural network and a novel recurrent

neural network classifier. Table 2 displays the mean

and standard deviation for the accuracy of the

Artificial Neural Networks and Novel Recurrent

Neural Networks algorithms, which stood at 94%

with a deviation of 1.79858, and 91% with a deviation

of 1.58348, respectively. The Artificial Neural

Networks demonstrated a smaller standard error of

.50074 in comparison to the Novel Recurrent Neural

Networks. The independent sample t-tests indicate a

significant difference in accuracy between the two

Comparing Novel Recurrent Neural Networks with Artiﬁcial Neural Networks for Predicting Mental Depression from Online Video Gaming

129

Table 3: The independent sample test revealed a substantial variation in accuracy among the suggested two stages and the

standard single stage. Since P=0.000 and p<0.05, there is a substantial variation between the two methods.

Independent Sample Test

Levene’s Test for Equality of Variances

T-test for Equality of Means

Sig.

Sig. (2-

tailed)

Mean Difference

Std. Error

Differences

95% Confidence Interval

of the Difference

Lower

Upper

Accuracy

Equal

Variances

assumed

.13

722

5.06

0.0

3.83

.75

2.24

5.4

Equal

Variances

not assumed

5.06

17.71

0.0

3.83

.75

2.24

5.4

Figure 1: Mean accuracy comparison of novel recurrent neural network (94%) and artificial neural network (91%) on online

gaming mental disorder identification. X-axis represents novel recurrent neural network and artificial neural network; Y-axis

represents mean accuracy ± 2 SD.

techniques. Table 3 underscores that the independent

sample test showcased a notable difference in

accuracy between the proposed two-stage approach

and the conventional single stage. Given that P=

0.000 (p<0.05), there exists a significant difference

between the two techniques.

Figure 1 depicts the mean accuracy results derived

from the artificial neural network method and the

proposed input relative to the selected input. The title

of the graph reads: "Diagnosing mental illnesses in

online gamers: A comparison of the mean accuracy

of artificial neural networks versus innovative

recurrent neural networks". The proposed technique

achieved a mean accuracy of 94%, surpassing the

91% accuracy of the artificial neural network. The X-

axis signifies accuracy while the Y-axis illustrates

mean accuracy ± 2SD.

4 DISCUSSION

From the tests, it's evident that the Novel Recurrent

Neural Networks algorithm outperformed the

Artificial Neural Networks. Accuracy was measured

using the SPSS tool for both the Novel Recurrent

Neural Networks and the Artificial Neural Networks.

The results for mental depression or disability in the

dataset indicated that the Novel Recurrent Neural

Network achieved an accuracy of 94%, which is

better than the 91% achieved by the Artificial Neural

Networks.

The focus of this study is the exploration of the

correlation between young adults' use of social media

and their mental well-being. Recent studies have

identified a connection between increased social

media engagement and declining mental health. This

link is particularly alarming given that young adults,

the most frequent users of social media, face an

unusually high risk of mental health challenges. This

dissertation delves into both individual and societal

theories that could elucidate the still enigmatic

relationship between social media engagement and

mental health. It examines the impact of social media

on personal relationships, inappropriate behaviour,

sleep disruption due to exposure to blue light, the

influence of social media on children, and the

repercussions of sedentary behaviour on mental well-

being.

AI4IoT 2023 - First International Conference on Artiﬁcial Intelligence for Internet of things (AI4IOT): Accelerating Innovation in Industry

and Consumer Electronics

130

Online gaming is associated with an elevated risk

of mental disorders, including depression. Both

Novel Recurrent Neural Networks and Artificial

Neural Networks are recent subjects of research as

promising tools for predicting and addressing the

adverse effects of online gaming on mental health

(Paulus et al. 2018). One potential application of

Artificial Neural Networks and Novel Recurrent

Neural Networks in this realm is to predict a gamer's

probability of developing depression based on their

gaming patterns and other factors (Biolcati, Pupi, and

Mancini 2021). The Novel Recurrent Neural

Networks, a subtype of Artificial Neural Networks, is

adept at analysing a player's in-game behaviour over

time since it excels in handling sequential data

(Hussain and Griffiths 2009).

Another possible application of Artificial Neural

Networks and Novel Recurrent Neural Networks in

the domain of online gaming and mental health is to

identify early indicators of potential problems

(Hussain and Griffiths 2009; Mancini, Imperato, and

Sibilla 2019; Jung, Yi, and JeongDongJin 2018). In

summary, Artificial Neural Networks and Novel

Recurrent Neural Networks possess the unique

capability to serve as potent tools for predicting and

mitigating the negative effects of online gaming on

mental health. However, further research is vital to

fully understand their potential and limitations in this

context (Jung, Yi, and JeongDongJin 2018).

Moreover, due to lockdowns and the subsequent

disruptions to work and education, individuals might

have excess leisure time, or they might find

themselves more easily distracted by online gaming

while working from home. Additionally, diverse

connections were observed between gamers'

motivations for playing and their choice of game

genres in relation to their psychological well-being.

Notably, those motivated by distraction and action

game enthusiasts displayed the most pronounced

effects. Further studies are essential to ascertain

whether these threats to mental health are caused by

or a consequence of video gaming.

5 CONCLUSION

The study at hand sought to utilise cutting-edge

machine learning methods to forecast mental

depression among online video game players. The

findings revealed that the Novel Recurrent Neural

Network algorithm notched an accuracy of 94%,

while its counterpart, the Artificial Neural Network

algorithm, achieved a slightly lower rate of 91%. In

juxtaposing the two, the Novel Recurrent Neural

Network algorithm exhibited superior performance in

the accuracy domain over the Artificial Neural

Network algorithm, with a mean accuracy difference

of 3.83500 between them. Delving deeper into the

research paper, it was accentuated that the Novel

Recurrent Neural Network algorithm, in the context

of predicting mental depression amongst online

gamers, surpasses the Artificial Neural Network

algorithm. Such results underscore the significance of

progressive machine learning methods, with special

emphasis on the Novel Recurrent Neural Network

algorithm, as promising tools for pinpointing and

addressing mental health challenges tied to online

gaming.

REFERENCES

AS, Vickram, Raja Das, Srinivas MS, Kamini A. Rao, and

Sridharan TB. "Prediction of Zn concentration in

human seminal plasma of Normospermia samples by

Artificial Neural Networks (ANN)." Journal of assisted

reproduction and genetics 30 (2013): 453-459.

Biolcati, Roberta, Virginia Pupi, and Giacomo Mancini.

2021. “Massively Multiplayer Online Role-Playing

Game (MMORPG) Player Profiles: Exploring Player’s

Motives Predicting Internet Addiction Disorder.”

International Journal of High Risk Behaviors and

Addiction. https://doi.org/10.5812/ijhrba.107530.

Blasi, Maria Di, Maria Di Blasi, Alessandro Giardina,

Cecilia Giordano, Gianluca Lo Coco, Crispino Tosto,

Joel Billieux, and Adriano Schimmenti. 2019.

“Problematic Video Game Use as an Emotional Coping

Strategy: Evidence from a Sample of MMORPG

Gamers.” Journal of Behavioral Addictions.

https://doi.org/10.1556/2006.8.2019.02.

Bonnaire, Céline, and Darlèn Baptista. 2019. “Internet

Gaming Disorder in Male and Female Young Adults:

The Role of Alexithymia, Depression, Anxiety and

Gaming Type.” Psychiatry Research 272 (February):

521–30.

Dias, Lucas Pfeiffer Salomão, Jorge Luis Victória Barbosa,

and Henrique Damasceno Vianna. 2018. “Gamification

and Serious Games in Depression Care: A Systematic

Mapping Study.” Telematics and Informatics.

https://doi.org/10.1016/j.tele.2017.11.002.

Fazeli, Sara, Isa Mohammadi Zeidi, Chung-Ying Lin,

Peyman Namdar, Mark D. Griffiths, Daniel Kwasi

Ahorsu, and Amir H. Pakpour. 2020. “Depression,

Anxiety, and Stress Mediate the Associations between

Internet Gaming Disorder, Insomnia, and Quality of

Life during the COVID-19 Outbreak.” Addictive

Behaviors Reports 12 (December): 100307.

Gackenbach, Jayne, and Johnathan Bown. 2017.

Boundaries of Self and Reality Online: Implications of

Digitally Constructed Realities. Academic Press.

G. Ramkumar and M. Manikandan, "Uncompressed digital

video watermarking using stationary wavelet

Comparing Novel Recurrent Neural Networks with Artiﬁcial Neural Networks for Predicting Mental Depression from Online Video Gaming

131

transform," 2014 IEEE International Conference on

Advanced Communications, Control and Computing

Technologies, Ramanathapuram, India, 2014, pp. 1252-

1258, doi: 10.1109/ICACCCT.2014.7019299.

Hussain, Zaheer, and Mark D. Griffiths. 2009. “Excessive

Use of Massively Multi-Player Online Role-Playing

Games: A Pilot Study.” International Journal of Mental

Health and Addiction. https://doi.org/10.1007/s11469-

009-9202-8.

Irie, Tomonari, Hiroki Shinkawa, Masanori Tanaka, and

Kengo Yokomitsu. 2022. “Online-Gaming and Mental

Health: Loot Boxes and in-Game Purchases Are

Related to Problematic Online Gaming and Depression

in Adolescents.” Current Psychology.

https://doi.org/10.1007/s12144-022-03157-0.

Jung, Kyeo-Woon, Inhyae Yi, and JeongDongJin. 2018.

“Effect of Gaming Motivation on Internet Gaming

Addiction in Massively Multiplayer Online Role

Playing Game (MMORPG) Users: Mediating Effects of

In-Game Behavior.” Korean Journal of Health

Psychology.

https://doi.org/10.17315/kjhp.2018.23.2.013.

Kishore Kumar, M. Aeri, A. Grover, J. Agarwal, P. Kumar,

and T. Raghu, “Secured supply chain management

system for fisheries through IoT,” Meas. Sensors, vol.

25, no. August 2022, p. 100632, 2023, doi:

10.1016/j.measen.2022.100632.

Mamun, Mohammed A., Irfan Ullah, Norina Usman, and

Mark D. Griffiths. 2022. “PUBG-Related Suicides

during the COVID-19 Pandemic: Three Cases from

Pakistan.” Perspectives in Psychiatric Care 58 (2): 877–

79.

Mancini, Tiziana, Chiara Imperato, and Federica Sibilla.

2019. “Does Avatar’s Character and Emotional Bond

Expose to Gaming Addiction? Two Studies on Virtual

Self-Discrepancy, Avatar Identification and Gaming

Addiction in Massively Multiplayer Online Role-

Playing Game Players.” Computers in Human

Behavior. https://doi.org/10.1016/j.chb.2018.11.007.

Maruta, N., O. Nazarchuk, and M. Denysenko. 2015. “The

Typology of Depressions with Comorbid Other Mental

Disorders.” European Psychiatry.

https://doi.org/10.1016/s0924-9338(15)32018-6.

Mun, Il Bong, and Seyoung Lee. 2021. “The Influence of

Parents’ Depression on Children’s Online Gaming

Addiction: Testing the Mediating Effects of Intrusive

Parenting and Social Motivation on Children’s Online

Gaming Behavior.” Current Psychology.

https://doi.org/10.1007/s12144-021-01854-w.

Paulus, Frank W., Susanne Ohmann, Alexander von

Gontard, and Christian Popow. 2018. “Internet Gaming

Disorder in Children and Adolescents: A Systematic

Review.” Developmental Medicine and Child

Neurology 60 (7): 645–59.

V. P. Parandhaman, "An Automated Efficient and Robust

Scheme in Payment Protocol Using the Internet of

Things," 2023 Eighth International Conference on

Science Technology Engineering and Mathematics

(ICONSTEM), Chennai, India, 2023, pp. 1-5, doi:

10.1109/ICONSTEM56934.2023.10142797.

Yuan, Kai, Ping Cheng, Tao Dong, Yanzhi Bi, Lihong

Xing, Dahua Yu, Limei Zhao, et al. 2013. “Cortical

Thickness Abnormalities in Late Adolescence with

Online Gaming Addiction.” PloS One 8 (1): e53055.

Yuan, Kai, Chen Wang Jin, Ping Cheng, Xuejuan Yang,

Tao Dong, Yanzhi Bi, Lihong Xing, et al. 2013.

“Amplitude of Low Frequency Fluctuation

Abnormalities in Adolescents with Online Gaming

Addiction.” PloS One 8 (11): e78708.

AI4IoT 2023 - First International Conference on Artiﬁcial Intelligence for Internet of things (AI4IOT): Accelerating Innovation in Industry

and Consumer Electronics

132