Fuzzy Expert System For The Selection Of Tourist Hotels

Transcription

1 INRNAIONAL JOURNAL OF SCINIFIC & CHNOGY RSARCH VOLUM, ISSU 09, SPMBR 06 ISSN Fuzzy pert System For he Selecton Of ourst Hotels PAL SINGH Abstract : In the present work, a smple and very effectve mathematcal model s desgned for tourst hotels of LVL. Locaton of hotels, buldng structure of hotels, qualty of hotels, feedback of hotels and advertsement of hotels are as nput factors. rapezodal membershp functon and trangular membershp functon are used for fuzzfcaton process and defuzzfcaton s done by COG technque. he fuzzy logc has been utlzed n several dfferent approaches to modelng the selecton of tourst hotels process. hs model addressed the hotel of LVL and ths model concludes that the hotel s LVL wth degree of precson. %. Key ords- Fuzzy logc, Fuzzy valdaton epert system, Lngustc varables, Root sum square (RSS). Introducton: In order to present a model of selecton of hotels, where the epectatons are fulfll for the arrval tourst. here s a huge problem n hotel ndustry to face the problems n selecton of tourst hotels. In ths study, we attempt to fnd a sutable response to ths research need by a mathematcal model, usng fuzzy epert system. herefore, hotel accommodatons are formulated by approprate fuzzy logc system. In ths order, we have to select a prme locaton for the hotels. he locaton s the man ssue n the hotel ndustry. Here, we have to try to solve the problems of selecton of tourst hotels by usng some mportant parameters. Many methods are developed n selecton of tourst hotels; almost evaluaton method has ts strong ponts and ssues for dfferent stuatons. In ths order, we wll use fuzzy logc to solve the problems of selecton of tourst hotels because hotels are man mportant part for postve thnkng n the feld of toursm. Good hotel locaton cannot only help ncrease market share and proftablty but can also enhance the convenence of customer lodgng because establshng a fne locaton wll shorten the payoff perod for fed captal nvestment.[9] Accurate forecasts of tourst arrvals and study of the tourst arrval patterns are the toursm related ndustres to formulate effcent and effectve strateges on mantanng and boostng toursm ndustry n a country.[] In order to develop the fuzzy decson makng model for selecton of tourst hotels. e wll focus on some parameters, whch wll help us n selecton of tourst hotels. In ths sesson, we use locaton of hotels, buldng structure of hotels, qualty of hotels, and feedback of hotels and advertsement of hotels. Here we have three level hotels for the tourst and try to control ths problem n a better way. A tourst hotel has a crucal role to developng and underdeveloped countres. Accurate forecasts on tourst hotels demand and study on the pattern of tourst demand from varous orgns are essental for the toursm. Selecton of tourst hotels depends upon the tourst s choces. Good qualty and good busness are the man demand n the tourst hotel ndustry. Gopal Sngh, Integrated Academy of Management and echnology, Ghazabad, Inda, PH , - mal: maths.gopal@hotmal.com In ths order, we can gather that nfluental factors for hotels to acheve success and reputaton, brandng of tourst hotels depends upon the management of the hotels. Here, each hotel requres attractng the tourst by ther good qualtes, good locaton, good atmosphere and good reputaton etc.. ourst hotels overvew: In order to present a model of selecton of hotels, where the epectatons are fulfll for the arrval tourst. here s a huge problem n hotel ndustry to face the problems n selecton of tourst hotels. In ths order, we have to select a prme locaton for the hotels. he locaton s the man ssue n the hotel ndustry and appromately 80 % of all cases of the tourst are related to hotels. very tourst have requrements that, how they choose ther level hotels. In a busness envronment, tourst hotels are good busness n the future. ourst demand forecastng focuses on tradtonal tme seres n tourst hotels. A tourst hotel has a crucal role to developng and underdeveloped countres. Accurate forecasts on tourst hotels demand and study on the pattern of tourst demand from varous orgns are essental for the toursm. Selecton of tourst hotels depends upon the tourst s choces. Good qualty and good busness are the man demand n the tourst hotel ndustry. In ths order, we can gather that nfluental factors for hotels to acheve success and reputaton, brandng of tourst hotels depends upon the management of the hotels. Here, each hotel requres to attract the tourst by ther good qualtes, good locaton, good atmosphere and good reputaton etc.. FUZZY GIC he concept of fuzzy sets was ntroduced n 96 by Lotf Zadeh (Zadeh L.A., 96) as a means of representng vagueness n applcatons. Fuzzy logc s the logc correspondng to fuzzy sets. In classcal two-valued logc, or Boolean logc or bnary logc, a proposton s ether true or false. he only permtted membershp values are 0 or. very tem n the unverse of dscourse s ether a full member of the set or not a member at all. wo valued logc works well for problems whch are lnear and systems that can be modeled precsely and t has proved to be effectve n solvng such problems. In multvalued logc, a proposton may be true, false or have an ntermedate truth value. he set of truth values s assumed to be evenly dvded over the nterval [0, ]. In fuzzy logc, the membershp functon can have values rangng from 0 to. IJSR 06

2 INRNAIONAL JOURNAL OF SCINIFIC & CHNOGY RSARCH VOLUM, ISSU 09, SPMBR 06 ISSN Modelng process decoraton, feedback s average or low and advertsement s ecellent or good then, ths s level hotel. Level : Hotel s establshed n cty and we are nearby to man tourst place, food s on demand, servce s low, rooms have medum qualty, buldng s old and good decoraton or buldng s new and good decoraton, feedback s low and advertsement s good or ecellent then, ths s level hotel.. Locaton of hotels: Fgure. Fuzzy Valdaton pert System. ypes of tourst hotels Here, we wll use fve nput parameters and try to fnd the level of hotel n the output. L (.) Locaton of hotels: Locaton of hotel s dvded n four parts: Low: Hotel s establshed n town area or nearby to cty. Average: Hotel s establshed n cty but we are far from man tourst place. Good: Hotel s establshed n cty and we are nearby to man tourst place. cellent: Hotel s establshed n cty and we have lot of choce to travel any tourst place. Qualty of hotels: Qualtes of hotels have many sub-parameters: Food, servce, room qualty, etc. akng all these subparameters n our mnd, we wll dvde qualty of hotels n four parts: Low: food s avalable by smple choces, servce s low, and rooms have smple qualty. Average: Food s on demand, servce s low, and rooms have medum qualty. Good: Food s on demand, servce s medum, and rooms have good qualty. cellent: Food s on demand, servce s very fast, and rooms have standard qualty. Buldng structure of hotels: Buldng structure s dvded n four parts: Low: Buldng s old and smple decoraton. Average: Buldng s old and good decoraton. Good: Buldng s new and good decoraton cellent: Buldng s new and ecellent decoraton. Feedback Of hotels: Feedback of hotels s dvded n four parts: low, average, good, ecellent. Advertsement of hotels: Advertsement of hotels s dvded n four parts: low, average, good, ecellent. Usng all these parameters, we wll decde the level of hotels: A 0 0 (.) G (.) μ( ) (.) Level : If hotel s establshed n cty but we are far from man tourst place or hotel s establshed n cty and we are nearby to man tourst place, food s avalable by smple choces, servce s low, rooms have smple qualty, buldng s old and good decoraton., feedback s low and advertsement s good then, ths s level hotel. Level : If hotel s establshed n cty but we are far from man tourst place, food s avalable by smple choces, servce s low, rooms have smple qualty or food s on demand, servce s low, rooms have medum qualty, buldng s old and good IJSR 06 Buldng structure of hotels: μ( )

3 INRNAIONAL JOURNAL OF SCINIFIC & CHNOGY RSARCH VOLUM, ISSU 09, SPMBR 06 ISSN L A (.6) (.7) G (.) (.) G (.8) μ( ) (.9). Feedback of the hotels X L 0 0 (.). Qualty of hotels X : A (.) L A (.0) (.) G (.6) IJSR

4 INRNAIONAL JOURNAL OF SCINIFIC & CHNOGY RSARCH VOLUM, ISSU 09, SPMBR 06 ISSN μ( ) (.7). efuzzfcaton he center of gravty (COG) method s used for defuzzfcaton process whch s the most popular technque and s wdely utlzed n actual applcatons. In ths method the weghted strengths of each output membershp functon are multpled by ther respectve output membershp functon center ponts and summed.fnally ths area s dvded by the sum of the weghted membershp functon strength and the result s taken as the crsp output. he COG method can be epressed as Output ata ma mn ma mn.. Advertsement of hotels X 0 ma{0, 0 0 L 0 ma{0, 0 0 (.8) A G 0 60 ma{0, 0 ma{0, (.9) ma{0, 70 ma{0, 70 (.0). Case-Study For the purpose of llustraton, we consder that the nature of the road usng four nputs vz locaton of hotels X, buldng structure of hotels X, qualty of hotels X, feedback of hotels X and advertsement of hotels X.. valuate the authentcty of the tourst hotels: he values of the nputs of the tourst hotels have to evaluated, X =, X =, X =, X =8, X =8 (say).. Fuzzfcaton of the crsp values of nputs: hrough the use of membershp functons defned for each fuzzy set for each lngustc varable. he degree of membershp of a crsp value n each fuzzy set s determned as follows: µ L ( ) =0, µ A ( ) =0.8, µ G ( ) =0., µ ( ) =0 (.) µ L ( ) =0.6, µ A ( ) =0., µ G ( ) =0, µ ( ) =0 (.) µ L ( ) =0, µ A ( ) =0.8, µ G ( ) =0., µ ( ) =0 (.) µ L ( ) =0.7, µ A ( ) =0., µ G ( ) =0, µ ( ) =0 (.) µ L ( ) =0, µ A ( ) =0, µ G ( ) =0., µ ( ) =0. (.6) Fre the rule bases that correspond to the nputs: Based on the value of fuzzy membershp functon values for the eample under consderaton, the followng rules apply: ma{0, 90 ma{0, 90 (. Rule No. If X X X X X G G G XC LLN he n LV L LV L μ( ) 9 G LV L 7 0 G O O G XC LLN XC LLN LV L LV L LV L IJSR

5 INRNAIONAL JOURNAL OF SCINIFIC & CHNOGY RSARCH VOLUM, ISSU 09, SPMBR 06 ISSN O O XC LLN LV L LV L. ecute the nference system e use Root Sum Square (RSS) method to combne the effects of all applcable rules. Root sum square method scales the functon at ther respectve magntudes and computes the fuzzy centrod of the composte area. hs method s more complcated mathematcally than other methods level L L L R Level R Level R Level 0. (0.) (0.) 0. References []. Zadeh, L. A. (968), Fuzzy algorthms, Info. & Ctl,, 9-0. []. Zadeh, L. A. (97), Outlne of a new approach to the analyss of comple system and decson process. SMC-,, 8-. []. Arbel, A. and Pzam, A. (977), Some determnants of urban hotel locaton: the tourst nclnatons, Journal of ravel Research,,, 8-. []. Ignzo, J. P. (99), "Introducton to pert Systems: he evelopment and Implementaton of Rule-Based pert Systems", McGraw-Hll, ISBN []. Gray,. S. and Lguor, S. C. (99), Hotel and motel management and operatons, New Jersey: Regents/Prentce Hall, -67. [6]. saur, S. H. and ze, G. H.(996),Multattrbute decson makng analyss for customer preference of tourst hotels, Journal of ravel & oursm Marketng,,, -69. [7]. Betzuen, A., Jmenez, M. and Rvas, J. A. (997), Actuaral mathematcs wth fuzzy parameters: An applcaton to collectve penson plans. Fuzzy conomc Revew, ( ), [8]. Urtasun, A. and Guterrez, I. (006), Hotel locaton n toursm ctes: Madrd , Annals of oursm Research,,, 8-0. Fgure:. Output of the decson of the epert system hs output shows that the hotel s LVL wth.% degree of precson. Concluson e use Center of gravty (COG) for defuzzfcaton. he defuzzfcaton of the data nto crsp output s accomplshed by combnng the results of the nference process. In the COG method, the weghted strengths of each output membershp functon are multpled by ther respectve output membershp functon center ponts and summed. Fnally, ths area s dvded by the sum of the weghted membershp functon strength and the result s taken as the crsp output. hs shows the crsp output belongs to LVL wth. % degree of precson. [9]. Chou,. Y., Hsu, C. L. and Chen, M. C. (008), A fuzzy mult- crtera decson model for nternatonal tourst hotels locaton selecton, Internatonal Journal of Hosptalty Management, 7, 9-0. [0]. Sharma, F. B., Sharma, A. and Sharma, N. (00), Fuzzy logc applcatons for traffc control an optmum and adaptve controllng applcaton. Internatonal journal of mergng echnologes (), -, []. Kumar, S. and Pathak, P. (00) Indcatve results of cancellaton of polces- Fuzzy approach. Proc. Natonal Conf. on Appl. Of math. to Sc. & ech., 7-. []. Kumar, S. and Pathak, P. (00) Fuzzy Based Bonus Malus System for Premum ecson In Car Insurance. Int. Revew Of Pure And Appled Maths., 6(), pp []. Hadavand,., Ghanbar, A., Shahanagh, K. and Naghneh, S. A. (0), ourst arrval forecastng by evolutonary fuzzy system, oursm management, IJSR

6 INRNAIONAL JOURNAL OF SCINIFIC & CHNOGY RSARCH VOLUM, ISSU 09, SPMBR 06 ISSN []. Rana,.H., Sarva, M.S., and Kumar, S. (0) A fuzzy logc based model for lfe nsurer underwrtng when nsurer s dabetc. urop. J. Appl. Sc., (), []. Kumar, S. and Jan. H. (0) Lfe nsurance underwrtng when nsurer s dabatc: A fuzzy approach. Int. J. Fuzzy Sys. & Rough Sys., (), -8. IJSR