Fuzzy Aggressive Behavior Assessment Of Toxic Players In Multiplayer .
Fuzzy aggressive behavior assessment oftoxic players in multiplayer online battle gamesGuilherme R. AndriguetoErnesto AraujoComputer Science ProgramUniversidade Anhembi Morumbi (UAM),São Paulo, BrazilBiomedical Engineering Post–Graduate ProgramUniversidade Anhembi Morumbi (UAM)Laureate International University (Laureate),Centro de Inovação, Tecnologia e Educação (CITE),Inteligência Artificial em Medicina e Saúde (IAMED),São José dos Campos, SP, Brazilernesto.araujo.br@gmail.comAbstract—A fuzzy system for assessing aggressive behavior ofplayers in electronic games of Multiplayer Online Battle type isproposed in this paper. Such an approach employs the fuzzylogic and the fuzzy set theory for evaluating and classifyingthe subjective feeling of aggressiveness, in particular, concerningtoxic players in this sort of competitive gaming. This paper aimsat achieving a meta analysis of the aggressiveness according tothe fundamental feelings within the psychology and psychiatric.Fear, happiness, anger and sadness are the input linguisticvariables that compose the input premise space mapped intothe aggressiveness outcome by using fuzzy IF–THEN rules. Sucha fuzzy aggressiveness assessment system is designed to stratifyand grade levels of emotional reactions that occur during thealso so–called electronic sports. The proposed approach comesto be an alternative for identifying and measuring altered temperduring, or after, the match.Index Terms—Toxic Players, Aggressive Behavior, Fuzzy Logic,Classification, Assessment, Multiplayer Online Battle GamesI. I NTRODUCTIONComputer games currently concern to the category ofelectronic sports that have been continuously growing. Simultaneously, it presents huge business potential, reflectingthe growing interest in competitive gaming. A pastime foradolescents and adults worldwide, this market deals with anaudience of approximately 2.1 billion players worldwide.Such a billionaire market and the success achieved inelectronic games have been prompting research on modelingthe profile and behavior of the users. A prominent researcharea, in particular, regards the aggressiveness among players,their relationship, whether games transform people, or evenwhether people use games drive their feelings, and so far [1].One style of game that became known worldwide for itsvery aggressive players is the Multiplayer Online Battle Arena(MOBA) [2]. This terminology is used to identify gameswhere two teams compete for an objective within a delimitedmap, each one with individual group strategies. In order toaccomplish complex team-based strategies, players take onvery specific roles within a team [3]. The multiplayer onlinebattle game developers, in general, bring about the aggressiveattitude of players during the game [4]. The behavior known asto troll in the gaming community refers to aggressive players978-1-7281-6932-3/20/ 31.00 2020 IEEEbehaving and being categorized by using verbal offenses aswell as to assign points within the games that induce tonegative actions leading the opposite team to lose the match[5].Nevertheless, aiming at the economic interest, MOBA developers are currently looking for a pleasant entertainmentenvironment for their non–aggressive players. Due to that,various actions to change the image of the game and theattitude of aggressive players have been taking, simultaneouslystimulating competitiveness and penalizing aggressive players.For instance, automatic conversion of aggressive into a nonaggressive phrase, temporary suspensions, loss of the accountof the aggressive player illustrate the strategy of producersand developers of games to reduce this sort of behavior inthe multiplayer online battle style. The question that comesup is if there would be a manner to measure such a subjectivefeeling of aggressiveness for classifying the player emotionalbehavior in the team-based multi-player game.In this paper a fuzzy system is designed to assess the fundamental emotional reactions and, thus, the aggressive behaviorthat a person outward when playing multiplayer online battleelectronic games. The system herein proposed is based on thefeeling wheel (Fig. 1) derived from the psychology along withFig. 1. Feeling Wheel representation of emotional reaction and behavior.
the fuzzy set theory and the fuzzy logic to design a meta–analysis model for aggressiveness assessment and analysistaking into account the feelings of fear, anger, happiness andsadness [6]–[8].The proposed approach aims at evaluating the influenceof those seminal emotions that lead to aggressive reactionsof such players. The output linguistic variable comprises thelevels of aggressiveness determined by the aggregation ofthese emotional reactions through fuzzy IF–THEN rules. Inso doing, the fuzzy player aggressiveness assessment systemis investigated as an alternative to stratify and to grade thediverse levels of aggressive reaction concerning players inMOBA electronic games.II. M ULTIPLAYER O NLINE BATTLE AGGRESSIVENESSA SSESSMENT S YSTEM BASED ON F UZZY L OGICOne of the most primitive and basic characteristics of allexisting animals is to ensure preservation. In the studies ofpsychology and psychiatry, to maintain and affirm the existence of the individual, the self-preservation can, in general,be associated with aggressive impulses. Characterizing thefeelings of an individual requires to assess the self-control,or lack of this condition, in social life. A feasible manner tocontextualize the emotional condition of and individual is thefeeling wheel, as shown in Fig. 1. In this sense, individuals’stillness, calmness, or aggression condition can be related tothe composition of the primary emotional conditions of thefeelings of fear, happiness, sadness, and anger.Constantly present while playing the game, the first inputlinguistic variable concerns fear that comes about at distinctinstants of time during the match, in different degrees of intensity both for each player. In general, when the player is goingto accomplish some new or different task, or is close to achievea goal, fear can be considered quite certain to be present. Inturn, anger is another emotion that can also be treated as part ofaggressiveness as various actions take place within the game.For instance, the constant loss of matches, troll of players,losing a goal, or actions that are not pleasing or are unexpectedcondition concerning the player, all of them lead to frustration.In contrast, happiness is generally associated with the feelingof reward. In the game context, it is the most prestigiousfeeling since the happiness can regard the player carried outa good move, accomplished a goal, or obtained the victory inthe match. Like other feelings, happiness can become harmfulif there is no control. Thus, a state of euphoria can affectthe player who may have a negative outcome in his behavior.Sadness is understood within the electronic game as anotherfeeling that occurs due to inherent negative actions presentin competition. Such a feeling is, however, characterized asnot being intense, but as being peaceful and smooth. Thisfeature can become peaceful if there is a sense of acceptanceof defeat. On the other hand, this feeling can trigger negativeunfolding resulting in explosive behavior if there is no control,most of the time concerning the frustration in achieving adesired condition within the match. Altogether, these feelingscan be considered a cornerstone to lead the player to a state ofFig. 2. Fuzzy system for assessment of aggressiveness in multiplayer onlinebattle game.emotional behavior from being under control or being out ofcontrol, respectively, related to calm or aggressive behavior.They comprise the meta–analysis risk factors related to theoutput linguistic variable concerning the aggressiveness level.It is worth mentioning, however, that happiness and sadnesscan be considered as opposing intensity in the same emotionaldimensional axis. Due to that, they are herein expressing in thesame domain, but with distinct directions. Further, the focusof the proposed system is to evaluate the temper status ofindividual in expressing aggressiveness, not being the objectiveto consider the broad expression of feelings. Due to that thelinguistic terms partitioning the output variable refer to levelsof classification according to aggressiveness during the match,and do not concern to psychological or psychiatric approach.In this sense, the proposed approach addresses the inputpremise space encompassing the fear, xFear , anger, xAnger ,happiness–sadness, xHappiness-Sadness , which compose the input three–dimensional Cartesian product XFear XAnger XHappiness-Sadness . These input linguistic variables are mappedinto the aggressiveness outcome, YAggressiveness , by using anonlinear input–output set of fuzzy IF–THEN inference rules,as depicted in Fig. 2.A. Fuzzy Input–Output Inference MappingThe fuzzy modeling employed in this paper uses the Mamdani type inference [9], prompt to imitate and to representknowledge concerning the experience of healthcare professionals. Characterized as a set of IF–THEN rules:Rj : IF hx1 is Mj1 (x1 )i AND . . .AND hxn is Mjn (xn )iTHEN hy is N i ,(1)the antecedent part, IF hpropositioni, defines the premise whilethe consequent part, THEN hpropositioni, refers to conclusion,both described by linguistic expressions in propositional form,P hx is M i. The j–th rule, j 1, 2, . . . , m, represents
the amount of rules, hxi is Mji (xi )i. The set of input fuzzypropositions, Pi i 1, . . . , n, where n is the number ofinput universe of discourse and represents the dimensionalityof the premises; and hy is N i, the inferred fuzzy proposition. The elements xi and y refer, respectively, to the i–thinput and the output concerning objects inserted in distinctclasses (sets) named universe of discourse, xi Xi andy Y , also assigned linguistic variables. The input vector,x [x1 , . . . , xn ]T , is related to the premises (antecedent ofthe rule) while the output, y, is associated to the conclusion(consequent of the rule). The linguistic expressions “AND”corresponds to the set operation, intersection, , logic operation, conjunction, , and Triangular norm operation, T–norm,t(x, y), . An operator : [0, 1]2 [0, 1] is called a T–norm if it is commutative, associative, monotonic and has 1as neutral element. When using the Mamdani fuzzy system,the T–norm is carried out by the minimum operation. Thedefuzzification operation is herein carried out by employingthe center of area. The elements Mi Xi and N Y arefuzzy sets and assigned linguistic terms, as well, partitioningthe respective universes of discourse.2) Anger: The second input linguistic variable refers to theanger. The set of linguistic terms TAnger {Annoyed, Furious}and their associated membership functions are distributed inthe universe of discourse with a range of XAnger [0, 10](Fig. 3(c)). In so doing, the membership functions becomeAnnoyedFuriousMAnger h5, 2.5, 1.11e 16i and MAnger h5, 2.5, 10i.3) Happiness–Sadness: The third input linguistic variablecorresponds to the compounding feelings of happiness andsadness, as illustrated in Fig. 3(b). Distinct of previous inputB. Input and Output Fuzzy SetsThe input fuzzy sets MjFear, and MjAnger , for jFear FearAngerjAnger 1, . . . , 2, meanwhile the input fuzzy setMjHapiness-Sadness and output fuzzy set NjAgressiveness , forAgressivenessHapiness-SadnessjHapiness-Sadness jAgressiveness 1, . . . , 4, have their membership functions defined according to the general description asfollows. Consider a membership function, µM : Xi [0, 1],defined upon an universe of discourse, Xi , to which is associated a set of terms T {M1 , M2 , M3 }; a linguistic termMj T , where c(Mj ) {x0 Xi µMj (x0 ) 1} ands(Mj ) {x0 Xi µMj (x0 ) 0}, respectively, denote thecore and support of Mj . In this paper, each linguistic termMj T is shaped according to a Bell membership function:(a) XFear : Fuzzy partition of the fear input variable.1µMij (xi ; a, b, c) 1 x ca2b(2)(b) XHappiness-Sadness : Fuzzy partition of the happinesssadness input variable.represented by ha, b, ci, where the slope is given by b/2a;a defines the width of the membership function, where alarger value creates a wider membership function, b defines theshape of the curve on either side of the central plateau, wherealarger value creates a more steep transition; and c defines thecenter of the membership function. The system is designed byemploying the Ruspini partitions.C. Input Linguistic Variables and Linguistic Terms1) Fear: The Fear input variable presents two partitions corresponding to the set of linguistic terms TFear {Rational, Irrational}, distributed into the universe of discourse in the range XFear [0, 10]. The set of terms forFearFearXFear is MRational) h5, 2.5, 1.11e 16i and MIrrational) h5, 2.5, 10i (Fig. 3(a)).(c) XAnger : Fuzzy partition of the fear input variable.Fig. 3. Input linguistic variables.
linguistic variables, such a universe of discourse is partitionedby four membership functions whose linguistic terms areTHappiness-Sadness {Dejected, Moody, Pleasant, Ecstatic}.DejectedTheir associated membership functions MHappiness-Sadness Moodyh3.32.5 10i, MHappiness-Sadness h3.32.5 3.3i,PleasantEcstaticMHappiness-Sadness h3.32.53.3i, MHappiness-Sadness h3.32.510i, are distributed in the universe of discoursewith a range of XHappiness-Sadness [0, 10].4) Output Diagnosing Variable: The output linguistic variable concerning the toxicity of multiplayer aggressivenessis also partitioned by employing Bell membership functionswhose membership functions are assigned the linguistic termsTAggressiveness {Alert, Normal, Tolerable, Unacceptable} distributed in a range of XAggressiveness [0, 10]. The membershipSeverityfunctions partition the universe of discourse as NAlert AggressivenessAggressivenessh3.32.5 10i, NNormal h3.32.5 3.3i, NTolerable Aggressivenessh3.32.53.3i, and NUnacceptable h3.32.510i (Fig. 4).The set of 2–4–2 linguistic terms that partition the input universes of discourse yields a set of 16 valid fuzzyregions in a two–dimensional input premise space, x [xFear , xHappiness-Sadness , xAnger ]T . Each region is mapped into thelinguistic terms that partition the output universe of discourserelated to the degree of emotional state of aggressiveness,according to a set of fuzzy IF–THEN inference rules.D. Fuzzy Multiplayer Online Battle Game AgressivenessRulesThe resulting Mamdani–based fuzzy multiplayer onlinebattle aggressiveness system1 is given as:R1 : IF hxFear is Lighti ANDhxHappiness-Sadness is LightihxAnger is LightiTHEN hAggressiveness is MildiR2 : IF hxFear is Lighti ANDhxHappiness-Sadness is ModerateihxAnger is LightiTHEN hAggressiveness is Mediumi.(3)R15 : IF hxFear is Severei ANDhxHappiness-Sadness is ModerateihxAnger is LightiTHEN hAggressiveness is SeriousiR16 : IF hxFear is Severei ANDhxHappiness-Sadness is SevereihxAnger is LightiTHEN hAggressiveness is SeriousiThe proposed system enables the gradual membership froman element to a class, yielding a smooth classification asshown in Fig. 5. It is worth mentioning that such a systemmay be employed with any shape of fuzzy sets available. Bellmembership functions are employed to illustrate the proposed1 Disclaimer: The fuzzy rules listed here should not be used in clinicaldiagnosis without consulting experienced physicians or psychologists.Fig. 4. Output linguistic variables for the fuzzy assessment: YAggressivenes :Fuzzy partition of the aggressiveness output variable.approach since it is taken into account that feelings presentcontinuous changing in temper, even when there are bursts ofemotional behavior.III. D ISCUSSION AND I LLUSTRATIVE E XAMPLESAt a glance, the Mamdani fuzzy multiplayer online battlegame aggressiveness IF–THEN system results in a nonlinearinput–output mapping, as depicted in Fig. 5. According tothe resulting MOBA fear-happiness-sadness-anger -based aggressiveness surface, the toxic player behavior comes aboutwhen these feelings active certain firing levels, regardless ofwhether they are common sense considered good or not. Suchan outcome surface is biased by the anger input linguisticvariable since it prompts a seminal influence in aggressiveness.The more intense and out of control each of these emotionalinput variables are, the worse is the aggressiveness, as it ispossible to observe in those graphics in Fig. 5. In this sense, asin any competition, the team should focus on the goal and noton any other element that affects the emotional state and drivesthe attention. In contrary, as the technical perspective hereinexposed, the analysis of those feelings detaches that when theyassume high values, the aggressiveness of the players can reachunacceptable levels.It is worth detaching the influence of the happiness-sadnessinput variable in the surfaces, taking into account the spacepremise of XHappiness-Sadness XFear and XHappiness-Sadness XAnger , respectively, illustrated in Fig. 5(a), and Fig. 5(b).In a glimpse, it is possible to observe the fading of theaggressiveness when the emotional reaction concerning happiness and sadness is in equilibrium, i.e., when the playeris not influenced by these feelings. The happiness-sadnessvariable imposes a notch in the response, contrary to mostof fuzzy (diagnosing, decision-making, assessment) systemswhere the surface does not present changes of this nature.Since the sadness is related to the negative signal and happiness is associated to positive signal, the emotional equilibriumcondition is around zero. Hence, achieving high negative and
well, that the happiness when out of control (high positivevalues) – represented by the linguistic term Ecstatic – presentsa worse aggressiveness than high negative values regardingintense sadness – represented by the linguistic term Dejected– prompting a non symmetrical surface outcome, as can beobserved in Fig. 5(a), and Fig. 5(b).(a)(b)Identifying the aggressiveness condition of the toxic playersby using the proposed fuzzy MOBA fear-happiness-sadnessanger system can be exemplified in Table I. Consider a player,P1 presenting a yet rational fear, xFear 4 meanwhile feelingpure ecstasy, xHappiness-Sadness 7.8, and a controlled angerfeeling mostly annoyed, xAnger 2. In this context, the systemoutcomes that the player scores yAggressiveness 4.3, achievingpredominantly the status of Alert from the monitoring perspective. The Normal and Tolerate classifications are also achieved,but firing with a lower intensity when compared to the higherdegree of activation obtained with the Alert membershipfunction. Taking into account the influence of sadness, contraryof the previous example, when happiness was assigned a highvalue, assume that a second player, P2 achieves the sameintensity, xHappiness-Sadness 7.8, keeping the other inputvariables unchanged. In this scenery, the aggressive reactionscores yAggressiveness 2.9. It is worth detaching in theseexamples that happiness and sadness differentiated herein onlyby the positive or negative signal of the measure, respectively.Further, the same intensity of happiness and sadness presentsP2P1, coherent with the yAggressivenessdistinct results, yAggressivenesssurface analysis as previously carried out. Such a conditionreflects the fact that sadness induce individuals to a state ofpassiveness meanwhile the happiness can lead an individualto activeness and, thus, being able to cause more conflict interms of harmful behavior. The proposed system also presentsdifferent stratifications. Although firing the same membershipfunctions in the output universe of discourse the degree of activation are characterized as µ Aggressiveness µ Aggressiveness MMNormalAlertAggressiveness for the second example, meanwhile the first oneMToleratethere is µ Aggressiveness µ Aggressiveness µ Aggressiveness .MMMAlertNormalTolerateµ(c)Fig. 5. Aggressiveness surfaces for multiplayer online battle game emotionalanalysis: Relationship between the happiness-sadness and the fear (5(a)); theanger and the happiness-sadness (5(b)); and the anger and the fear (5(c)) foraggressiveness assessment.high positive values directly worsen the aggressiveness in theoutput variable. Moreover, around zero the individual is notinfluenced by such emotional reaction such that keeping calmand focused would be the best strategy to be employed in gamecompetitions, whether online or not. Despite referring to equalimportance, these two feelings do not affect proportionallyand similarly the aggressiveness. It is worth detaching, asAfterward, three individuals presenting the same average measures of fear, xFear 6.6, and happiness-sadness,xHappiness-Sadness 4.3, but with distinct anger intensity aredescribed to represent the influence of this latter variableupon the aggressiveness assessment. The first individual,P3 , presents the anger intensity of xAnger 2; the second one, P4 , is characterized as xAnger 5.5; and thethird, associated to xAnger 9. The resulting outcomesare, respectively, yAggressiveness 3.6, yAggressiveness 5.2,and yAggressiveness 5.9. These individuals fire all themembership functions of the output universe of discourse,but differ by the degree of activation that each class assume. For instance, for P3 there is µ Aggressiveness MAlertAggressiveness µ Aggressiveness µ Aggressiveness ; meanwhileMMMNormalTolerateUnacceptablefor P4 , µ Aggressiveness µ Aggressiveness µ Aggressiveness MMMAlertNormalTolerateµ Aggressiveness ; and, finally, for P5 , there is µ Aggressiveness MMUnacceptableTolerateµ
TABLE II LLUSTRATIVE E XAMPLE OF FUZZY MULTIPLAYER ONLINE BATTLE GAME AGGRESIVENESS StratificationStratificationScore4Rational*, Irrational7.8Pleasant, Ecstatic*2Annoyed*, FuriousNormal, Alert*, Tolerable,4.3P24Rational*, Irrational-7.8Dejected*, Moody2Annoyed*, FuriousNormal*, Alert, Tolerable,2.9P36.6Rational, Irrational*4.3Pleasant2Annoyed*, FuriousNormal, Alert*, Tolerable, Unacceptable3.6P46.6Rational, Irrational*4.3Pleasant5.5Annoyed, Furious*Normal, Alert*, Tolerable**, Unacceptable5.2P56.6Rational, Irrational*4.3Pleasant9Annoyed*, FuriousNormal, Alert, Tolerable*, Unacceptable5.9P64.5Rational*, Irrational8.9Ecstatic9.4FuriousNormal, Alert, Tolerable**, Unacceptable*6.5P76.5Rational*, Irrational8.9Ecstatic9.4FuriousNormal, Alert, Tolerable, Unacceptable*7.3AggressivenessAlert µMAggressivenessNormal µMMembership function with the higher degree of activation.Aggressiveness .UnacceptableThe factthat the same membership functions are activated for theseexamples simultaneously that present different scores meansthat the proposed system enables to capture the approximatereasoning by flexing the perception of the subjectiveness ofsuch a complex variable.The influence of fear in the aggressiveness assessmentis illustrated onwards. Consider two individuals who arecharacterized as being furious, xAnger 9.4, and in ecstasy (happiness), xHappiness-Sadness 8.9. Although the rational fear, xFear 4.5, of one individual, P6 , is closeto the irrational fear, xFear 6.5, of the player P7 ,the quantitative aggressive measures assume distinct values,yAggressiveness 6.5, yAggressiveness 7.3, respectively. Likewiseprevious example, in which all the membership functionsare active, the qualitative aggressive measures are distinctby the degree of activation of the classes. In this sense,while P6 is described by µ Aggressiveness µ Aggressiveness MµMAggressivenessNormal MTolerateUnacceptableAggressiveness . In turn, the individMTolerateµ Aggressiveness µ Aggressiveness MMUnacceptableTolerateµ Aggressiveness . In this sense, when the fearMAlert AggressivenessMNormalual P7 presentsµAggressivenessStratification MFearScaleP1µFuzzy Aggresiveness AssessmentHappiness-Sadnessµchange from rational to irrational, concerning an increasingvariation of xFear 2, inflicts changes in the evaluation ofthe aggressive behavior from Tolerate to Unacceptable.As it is possible to notice, the proposed approach enablescapturing the subjective influence of more than one emotionalbehavior and reaction of fear, happiness, sadness, and angerrelated to the feeling wheel that, in turn, comes to be a singlemeasure, since the fuzzy system also scores this emotionalcondition.IV. C ONCLUSIONIn this paper, assessing the aggressiveness is carried outby employing the primary feelings of fear, happiness, angerand sadness that compose the felling wheel to representemotions. A set of fuzzy IF-THEN rules grade altered temperto support stratifying emotional reactions simultaneously thatsuits to capture the approximate reasoning by flexing theperception of the subjectiveness of such a complex variable. The resulting fuzzy aggressiveness assessment systemaddresses the behavior of very aggressive – i.e., toxic –players herein used for multiplayer online battle arena gameevaluation. Future work extends this fuzzy fear-happinesssadness-anger system to deal with experimental data afterbeing approved by the human ethical committee to carry outpractical assessment, also analyzing exogenous confoundingor modulation (scheduling) factors that can interfere in thisaggressiveness emotional analysis. The resulting fuzzy fearhappiness-sadness-anger aggressiveness assessment can be analternative for MOBA producers and developers to supportidentification toxic players as well as aggressive behaviorduring, or after, the match.R EFERENCES[1] Adrienne H. Ivory, Christine E. Kaestle, The Effects of Profanity inViolent Video Games on Players’ Hostile Expectations, AggressiveThoughts and Feelings, and Other Responses, Journal of Broadcasting& Electronic Media, v.57, n.2, pp. 224-241, 2013.[2] Justin W. Bonny, Lisa M. Castaneda, Impact of the Arrangement ofGame Information on Recall Performance of Multiplayer Online BattleArena Players, Applied Cognitive Psychology, v.30, pp. 664-671, 2016.[3] Adam S. Kahn, Dmitri Williams, We’re All in This (Game) Together:Transactive Memory Systems, Social Presence, and Team Structure inMultiplayer Online Battle Arenas, Communication Research, v.43, n.4,pp. 487-517, 2016.[4] Filip Nuyens, Jory Deleuze, Pierre Maurage, Mark D. Griffiths, DariaJ. Kuss, Joel Billieux, Impulsivity in Multiplayer Online Battle ArenaGamers: Preliminary Results on Experimental and Self-Report Measures, Journal of Behavioral Addictions, v.5, n.2, pp. 351-356, 2016.[5] Enric Bertran, Andres Chamarro, Videogamers of League of Legends:The role of passion in abusive use and in performance, Adicciones, v.28,n.1, pp. 28-34, 2016.[6] Gloria Willcox, The Feeling Wheel: A Tool for Expanding Awarenessof Emotions and Increasing Spontaneity and Intimacy, TransactionalAnalysis Journal, v. 12, n. 4, pp. 274-276, 1982.[7] L.A. Zadeh, Fuzzy Sets, Information and Control, vol. 8, 1965, pp. 338–353.[8] L.A. Zadeh, The concept of a linguistic variable and its application toapproximate reasoning, Information Sciences, vol. 8, no. 9, pp. 43–80,1975.[9] E. H. Mamdani, S. Assilan, An experiment in linguistic synthesis witha fuzzy logic controller, Intern. Journal of Man-Machine Studies, vol.7, pp. 1–13, 1975.
Index Terms—Toxic Players, Aggressive Behavior, Fuzzy Logic, Classification, Assessment, Multiplayer Online Battle Games I. INTRODUCTION Computer games currently concern to the category of electronic sports that have been continuously growing. Si-multaneously, it presents huge business potential, reflecting the growing interest in .
ing fuzzy sets, fuzzy logic, and fuzzy inference. Fuzzy rules play a key role in representing expert control/modeling knowledge and experience and in linking the input variables of fuzzy controllers/models to output variable (or variables). Two major types of fuzzy rules exist, namely, Mamdani fuzzy rules and Takagi-Sugeno (TS, for short) fuzzy .
fuzzy controller that uses an adaptive neuro-fuzzy inference system. Fuzzy Inference system (FIS) is a popular computing framework and is based on the concept of fuzzy set theories, fuzzy if and then rules, and fuzzy reasoning. 1.2 LITERATURE REVIEW: Implementation of fuzzy logic technology for the development of sophisticated
Different types of fuzzy sets [17] are defined in order to clear the vagueness of the existing problems. D.Dubois and H.Prade has defined fuzzy number as a fuzzy subset of real line [8]. In literature, many type of fuzzy numbers like triangular fuzzy number, trapezoidal fuzzy number, pentagonal fuzzy number,
Fuzzy Logic IJCAI2018 Tutorial 1. Crisp set vs. Fuzzy set A traditional crisp set A fuzzy set 2. . A possible fuzzy set short 10. Example II : Fuzzy set 0 1 5ft 11ins 7 ft height . Fuzzy logic begins by borrowing notions from crisp logic, just as
of fuzzy numbers are triangular and trapezoidal. Fuzzy numbers have a better capability of handling vagueness than the classical fuzzy set. Making use of the concept of fuzzy numbers, Chen and Hwang [9] developed fuzzy Technique for Order of Preference by Similarity to Ideal Solution (TOPSIS) based on trapezoidal fuzzy numbers.
ii. Fuzzy rule base: in the rule base, the if-then rules are fuzzy rules. iii. Fuzzy inference engine: produces a map of the fuzzy set in the space entering the fuzzy set and in the space leaving the fuzzy set, according to the rules if-then. iv. Defuzzification: making something nonfuzzy [Xia et al., 2007] (Figure 5). PROPOSED METHOD
2D Membership functions : Binary fuzzy relations (Binary) fuzzy relations are fuzzy sets A B which map each element in A B to a membership grade between 0 and 1 (both inclusive). Note that a membership function of a binary fuzzy relation can be depicted with a 3D plot. (, )xy P Important: Binary fuzzy relations are fuzzy sets with two dimensional
the traditional fuzzy c-means to a generalized model in convenience of application and research. 2.1 Fuzzy C-Means The basic idea of fuzzy c-means is to find a fuzzy pseudo-partition to minimize the cost function. A brief description is as follows: (1) In above formula, x i is the feature data to be clustered; m k is the center of each clus-ter; u
