Electron Commer Res DOI 10.1007/s10660-017-9259-6
Game theoretic approach of a novel decision policy for customers based on big data Shasha Liu1 • Bingjia Shao1 • Yuan Gao2,5,6 Su Hu3 • Yi Li4 • Weigui Zhou2
•
Springer Science+Business Media New York 2017
Abstract In recent days, big data based analysis in hotel industry become popular. Merchants are attracting clients using the accurate analysis of historic data and predicting the behavior of possible clients to perform proper marketing strategy. To study the principle of the game between clients and merchants, in this work, we propose a novel two-stage game theoretic approach of decision policy for clients when choosing the suitable hotel to stay among many candidates, the merchants will provide a non-cooperative game strategy to attract the attention of potential clients. Analysis of the non-cooperative game method based on big data has been given. Simulation results indicate that, by using our proposed novel method, the average price for clients to choose a satisfied hotel is reduced and the successful rate of stay is increased for merchants, which will bring the expected income to a higher level because of the sticky phenomena of users.
& Bingjia Shao
[email protected] & Yuan Gao
[email protected] 1
School of Economics and Business Administration, Chongqing University, No.174 Shazheng Street, Shapingba District, Chongqing 400030, China
2
Xichang Satellite Launch Center, No. 52 Changanbei Rd., Xichang 615000, Sichuan, China
3
University of Electronic Science and Technology of China, Chengdu 611731, Sichuan, China
4
The High School Affiliated to Renmin University of China, No. 37 Zhongguancun Rd., Beijing 100080, China
5
China Defense Science and Technology Information Center, No.26 Fucheng Rd., Beijing 100142, China
6
Department of Electronic Engineering, Tsinghua University, Beijing 100084, China
123
S. Liu et al.
Keywords Game theory Big data Accurate prediction Non-cooperative game Decision policy
1 Introduction As highly regarded by commercial market, big data has become popular in many areas, where historical data could help make accurate consideration of economic phenomena. Hotel industry is one possible area to use big data [1] analysis, where advertisements are required to attract users’ attention [2] and provide possible increase of income. Recently, many hotel companies such as IHG and SPG are becoming popular all over the world, for its good quality and influence of brand, people are enjoying the quality of trip by utilizing the information and service provided by hotel companies and member plan. Also there are some companies that gather all the hotel information together and become agents for customers to earn the difference from price, such as Ctrip Co.Ltd and Booking.com. The asymmetric information between customers and hotel companies make the process as a game, so hotels are trying their best to design advertisement, special offer and member plan, etc. to squeeze money from customers. On the contrary, the customers are seeking the suitable hotel to stay, considering the influence of price, position, condition, service, etc. Considering the game between hotel companies and customers, a proper way to help the customers to decide the suitable hotel is in great demand. So the game theoretic way to play a non-cooperative game for clients is one possible area to tackle. 1.1 Related work Researchers are trying to identify the exact the way to help users to decide the proper hotel to stay, using some economics methods, such as game theory, decision theory, etc. The effective way to tackle the game between clients and hotel is to analyze the data obtained from privacy information and historic record. In [3], the author studied the big data based geotargeting method to decide the target using the geo-information, however, the geo-information is only obtained from the mobile terminals, such as integrated GPS module, such information is incomplete, so the accuracy of the method is limited. Through some analysis of the given market [4, 5], the hotel could run professional marketing policy [6] [7] to attract customers, e.g. the precise advertisement. Bone, etc. presents the study of peer-to-peer problem solving method while the customers could participate [8], this is the exact way that the customers and the hotel companies could cooperate. Thus, the hotel companies want to improve the occupancy rate at any given time, this is a hard and impossible problem for most of the companies, because the occupancy rate is relying on many influence factors, such as season for tourist, climate, price, etc. Then in [9] and [10], the authors provide possible solution using big data and mining technology, the hotel companies collect intelligence from other agencies and analyze the huge amount of data, then decide the policy to attract the suitable customer group, such as High Net Worth Clients (HNWC), Economic Clients (EC). In
123
Game theoretic approach of a novel decision policy for…
[11], the authors give their service oriented system design, by introducing the importance of service, the proper evaluation mechanism is raised to grade the different types of service [12], but the method is not quite complete for the service evaluation in hotel industry is relying on the independent third-party involvement. In [13] and [14], big data based evaluation method is introduced, which could be used to obtain necessary information from network side or submitted by customers as personal information. Then in [15] and [16], the authors improve the grade methods using big data analysis, some experimental results are discussed, but no practical algorithms [17] are given to make the method work. In [18] and [19], big data prediction method is proposed; the big data based prediction method is used to predict the upcoming load of check-in users. Then they arrange the suitable room for customers. In the game between hotel and customers, the accuracy of prediction is directly affected the sales promotion, so in [20], the real example of prediction using big data is given, where the results indicate that big data prediction is limited by the amount of data and the quality of data obtained from customers. So in [21, 22] and [23], the authors tried to discover the proper way to tackle the balance between sales and customers through game theory using big data. However, they did not provide the suitable method to implement the game. 1.2 Contributions of this paper In this paper, considering the demand from both customers and hotel industry, we study the game between hotel and customers and then propose the decision policy for users to decide the proper hotel to live among some choices. The main contribution of this work is summarized as follows. First, we analyse the need of customers and hotel operators, and point out the decision factors that may affect the policy adopted by customers, such as position (factor), environment, price and special offer; Second, we model the process of in-cooperate game between customers and hotels, the big data based analysis is given to ensure the prediction of our method; Third, we propose the practical decision policy to decide the proper way to choose the hotel, the customer could obtain the satisfactory at the most, and simulation result verified the discussion in this part. 1.3 Structure of the paper The rest of the paper is organized as follows. In part 2, we propose the game model between customers and hotels using big data, analysis of influence factors is also given in this part. In part 3, we propose the novel decision policy based on big data analysis, the algorithm is analyzed in this part. In part 4, simulation and analysis are presented. Finally, conclusion and outlook are given.
123
S. Liu et al.
2 Game model As mentioned above, game between customers and hotel operators are popular in recent researches. The decision policy of user is based on the objective information obtained from network or hotels, such as price, condition. Nevertheless, in fact, the information obtained by customers is not neutral, so the game is called asymmetric game. In Fig. 1, we describe the basic model of the game. For every possible customer that will choose a hotel to stay, there are multiple candidate hotels tagged from one to N. For any selected hotel, there are three possible operations for a customer: Check-in without bargain: this status means that the customer accepts the price of the room and choose to stay without any hesitation. Check-in with bargain: the customer wants to stay in the hotel but the price is not satisfactory, so a bargain is needed. If the hotel cut down the price, the customer would accept the new price and choose to stay.
Candidate Hotel
Bargain Status
Hotel N
Check-In No Bargain
Hotel N-1
Check-In With Bargain
Hotel N-2 Not Check-In With Bargain
Hotel 1
Fig. 1 Game between customers and hotels. For every user, candidate hotels are ready to be selected, there are three possible statuses: check-in without bargain, check-in with bargain and bargain but not check-in
123
Game theoretic approach of a novel decision policy for…
Not Check-in with bargain: the customer tried to bargain the price, but no balance has been reached, so the customer decided to choose another hotel to stay. However, the price is not the only influence factor that affect the decision of customers, so the compound issues will be discussed in the following sub-section. 2.1 Influencing factor The decision factors for customers are complex; the proper model to mix the factor could help solve the problem in an effective way. Considering the multiple influence factors that will affect the decision, we will model some of the influencing factors and model the price function to make the proposed method work well. 2.1.1 Price We use M [ N[0, ??) to define this variable, this is the basic and famous factor that may affect the decision of customers, for different customers, the tolerance of M is quite different. Mi is the price for hotel i [ I, respectively. 2.1.2 Distance We define Hi = (xi, yi) the coordinate of the hotel i from target customer, and the pffiffiffiffiffiffiffiffiffiffiffiffiffiffi distance Di ¼ x2i þ y2i 2 N½0; Dmax ; Dmax 2 N þ means the maximum accepted distance from the target customer with coordinate(0,0). 2.1.3 Environment We define E = {1, 2, 3, 4, 5}, for each candidate hotel i; the user could mark the environment factor Ei to describe the classification of the hotel. Level 1 means the lowest and level 5 is the highest. 2.1.4 Policy For there are some hotels that will provide membership card and award plan for Favored Client Discount, we model the factor called policy, where Li [ N[0,1] indicates the influence of award plan for hotel i. 2.1.5 Bargain policy The customer will have a policy to perform bargain, so the solution is defined using a [ N[0,1], it is a factor that describe the will of customer that want to have the bargain with the hotel with price. The decision whether the customer choose to stay, is defined using the utility functionUs(a), where it is a complex problem and defined using Eq. (1).
123
S. Liu et al.
1þa Ps Us ðaÞ ¼ d aPi 2
ð1Þ
where d [ N[0,1) is the control factor to decide the scale of the function, Pi is the normalized price to check-in with hotel i and Ps is the average normalized discount that the customer abandoned, the definition of the two functions is listed in Eqs. (2) and (3). E i Mi Pi ¼ Li log2 1 þ ð2Þ Di P EM Lj log2 1 þ Dj j j Ps ¼
j2I; j6¼i
I1
ð3Þ
In Eq. (2) and (3), EiMi means the price with environment factor, EiMi/Di is the normalized price compared to distance. Ps is the average price that other candidate hotel could provide considering all possible influence factors. Define Ubound the accepted bound that the customer will decide to check in, so obviously, if Us(a) [ Ubound, the customer will choose to stay without any hesitation, if - Ubound B Us(a) B Ubound, then the customer hesitate to check in, and if Us(a) \ - Ubound, then there must exists some cost-effective hotels to stay, so the best choice for customer at this time is to find some other hotels. We can infer that, (1 ? a)P/2 is approximated to the accounting cost, which represent the price of choosing hotel i. If no bargain is introduced, the customer will choose hotel i with the given price Mi. 2.2 Game between customers and hotels For any given observation period, the customer will choose one hotel to live in. The process of bargain may contain multi-stages, just like several rounds of negotiation in commercial market. The target of the game is to find out a balance between hotels and customers, so the multi-stage bargain is introduced [24, 25]. To make the analysis simple, we only consider the two-stage bargain model between customer and hotels. At the beginning of the first round bargain, the customer will choose d1 as policy to decrease the utility function Us(a) to increase the faith, when the price is within the confidence interval, the second round bargain will start, the customer will choose d2 as new policy, then the utility function Us(a) will change and reach a new bound, the distribution of the faith is no longer a uniform distribution. In Fig. 2, the sequential decision and possible solution are listed. The customer will make decision that whether the bargain is accepted. Because there exist two chances to perform bargain, the customer may reject the price even when the utility function is within its confidential bound, and expect the second round bargain. Here we assume that the price and the policy will change in different round of bargain, so variation is obtained and the bargain is effective. The customer will update the expected price and utility function in every round of bargain, then the
123
Game theoretic approach of a novel decision policy for… Fig. 2 Two-stage bargain game between customers and hotels. In the two stages, if the customer accepts the stage 1 price, no more bargain is needed; elsewise the second round bargain is introduced with different d
Stage 1 Reject Price
Accept Price
Stage 2
Accept Price
Reject Price
hotel and the customer will reach the conclusion. The faith of any customer is listed below: Faith: l1(M) [ U[0, j] is the faith to the price in the first round bargain, l2(M|a1) is the updated faith when the customer reject the policy a1 in the second round. Assume j is large enough, define the upper bound of price: M ða1 ; a2 Þ ¼
2Ms ða1 d2 a2 Þ Ps ½ð1 þ a1 Þ d2 ð1 þ a2 Þ
ð4Þ
3 Novel proposed game policy Because there are two possible bargain processes, we first give the policy in the first round. 3.1 Stage 1 bargain Lemma 3.1 The customer will reject the first round price when any of the conditions is satisfied: h i 1 Ms 1. M 2 Ps2að1þa ; j , and a1 [ a2 1Þ 2. 3.
M [ [M*(a1, a2), j], and d2a2 \ a1 \ a2 M [ [0, j], and a1 \ d2a2
123
S. Liu et al.
Elsewise the customer will accept a1. Proof In the two-stage bargain, the customer will choose different policy to obtain different utility: 1. 2. 3.
1 If the customer accept a1, then Us ¼ d1 a1 Ms 1þa 2 d1 MPs If the customer accept a2 and reject a1, then we have the utility function: 2 Us ¼ d1 d2 a2 Ms 1þa 2 d1 d2 MPs If the customer reject all, then Us = 0,
So it is clear that, when: d1 a1 M s
1 þ a1 1 þ a2 d1 MPs [ d1 d2 a2 Ms d1 d2 MPs 2 2
ð5Þ
Compared to accept a2 and reject a1, the customer prefer to accept a1. Similarly, we have: 1 þ a1 d1 MPs [ 0 ð6Þ d1 a1 M 2 It is obvious that the customer prefer to accept a1 compared to reject all the two options. When a1 [ d2a2, according to Eq. (5), we have 0 \ M \ M*(a1, a2), and if we consider Eq. (6), we have: 2a1 Ms 0 \M \min M ða1 ; a2 Þ; ð7Þ Ps ð 1 þ a 1 Þ Then the customer will accept a1. To conclude, we have two conditions listed below: 1. 2.
1 Ms When a1 [ a2, we have 0 \ M\ Ps2að1þa 1Þ When d2a2 \ a1 \ a2, we have 0 \ M \ M* (a1, a2)
It is clear that Eqs. (5) and (6) will not work together, so for any M [ [0, j], the customer will not accept a1. Lemma is proved. For any solution a1 in the first round, the policy of customer is expressed using the following expression: a2 ða1 Þ is the solution to the fixed-point equation about a2: j1 ða1 ; a2 Þ j ; 1 a2 ¼ min max ap ðj1 ða1 ; a2 ÞÞ; ; min 2Ms =Ps j1 ða1 ; a2 Þ 2Ms =Ps j ð8Þ where ap ðj1 ða1 ; a2 ÞÞ
123
sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2Ms ð2M Mbound Þ 1 ¼ Ps Mr ð2Ms =Ps j1 ða1 ; a2 Þ 2
ð9Þ
Game theoretic approach of a novel decision policy for…
and j1 ða1 ; a2 Þ ¼
2Ms ða1 d2 a2 Þ Ps ðð1 þ a1 Þ d2 ð1 þ a2 ÞÞ
ð10Þ
Then the optimal solution of the first round bargain is obtained using the convex optimization toolbox: 0 1 ðd1 ð1 a1 ÞMr M bound ÞP1 a1 ¼ arg max @ þ d1 d2 1 a2 ða1 Mr Mbound P2 A ð11Þ a1 2½0;1 þðd1 d2 1ÞMbound P3 In the above solution, definition of P1, P2, P3 are the probability of every policy, which are listed below, when d2 a2 a1 \a1 \a2 , the faith and the policy consist of a sequential equilibrium solution in stage 1. P1 ¼
j 1 ð a1 Þ j
2Ms a2 Ps ð1þa2 Þ
P2 ¼ P3 ¼
j1 ð a 1 Þ
ð12Þ
j
s a2 j Ps2M ð1þa2 Þ
j
3.2 Stage 2 bargain Then we continue to the stage 2 game of the customer like the analyze in stage 1. Lemma 3.2 The following faith and policy consist of infinite pairs of sequential equilibrium solution: j ; 1 ð13Þ a2 ¼ min max ap ; 0 ; min 2Ms =Ps j ap
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2Mr Mbound 1 ¼ 2 Mr
ð14Þ
And the parameters must satisfy: 1. 2. 3. 4.
a1 : any positive number that satisfy a1 \d2 a2 l2(M) = l2(M|a1): normal distribution l1(M) [ U[0, j] of customers @1(a1|M): the customer reject a1 2 @2(a2|M, a1): when d1 d2 a2 Ms 1þa 2 d1 d2 Ps M [ 0, the customer accept a2
123
S. Liu et al.
Proof Considering M [ [0, j] and a1 \ d2a2, because the solution a1 in stage 1 is far more small than that in stage 2, so the customer will not accept a1. Based on this observation, we have the optimal policy in stage 2: j ; 1 a2 ¼ min max ap ; 0 ; min ð15Þ 2M =P j s
s
rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 2M Mbound 1 ap ¼ 2 Mr
ð16Þ
Because a2 is independent to a1, and it is a constant when other system parameters are fixed, so when a1 is given, the policy is listed below: j 1. a2 ¼ min max ap ; 0 ; min 2Ms =P ; 1 ; constant s j 2. 3. 4.
l1(M) = l2(M|a1): normal distribution l1(M), l2(M) [ U[0, j] of customers @1(a1|M): the customer reject a1 2 @2(a2|M, a1): when d1 d2 a2 Ms 1þa 2 d1 d2 Ps M [ 0, the customer accept a2
In the above solution, the three variable P1, P2, P3 defined in Eq. (12) are listed below: P1 ¼ 0 2Ms a2 jPs 1 þ a2 2Ms a2 P3 ¼ 1 jPs ð1 þ a2 Þ
P2 ¼
ð17Þ
Then the optimal a1 2 N ½0; 1 will be the solution to maximize the following equation: 2a2 Ms Up ða1 Þ ¼ d1 d2 1 a2 Mr Mbound jPs ð1 þ a2 Þ 2a2 Ms þ ðd1 d2 1ÞMbound 1 jPs ð1 þ a2 Þ
ð18Þ
Because a1is not the variable in Eq. (18), so a1 2 N ½0; 1 could be any positive value, lemma is proved.
4 Simulation and analysis In this part, we present the simulation and analysis of the game between customers and hotels. Important to say, the big data is introduced in this part and the influence is added to our novel optimal game policy.
123
Game theoretic approach of a novel decision policy for…
6SDFHGRPDLQ +RWHO1
+RWHO71
ĂĂ ĂĂ
+RWHO +RWHO +RWHO
+RWHO
ĂĂ
+RWHO7
7LPHGRPDLQ Fig. 3 Big data structure obtained from service provider. The two-dimension data contains time and hotel information, which means, for every hotel, we have every day check-in price
4.1 Big data modeling We know that there are many data providers that could provide the historical data of the price. In part 2.1, we define the parameter policy: Li [ N[0,1], in simulation part, we will use this parameter to indicate the influence from historical data. Assume we could obtain historical price data from the third party service provider, with the data structure described in the Fig. 3. For every hotel, we collect the 365-day data to describe the trends of the price, so for every hotel i, define the normalized policy using the following equation: Li ðtÞ ¼
Mi ðtÞ=sumðMi ðtÞÞ P i2I Mi ðtÞ
ð19Þ
4.2 Simulation and analysis The simulation parameters are defined firstly. j = 1.5Ms/Ps is the upper bound of the variable M, d1 = 0.5, d2 = 0.7, this means that in the second stage, the customer will accept more bargain condition, Mbound = 1,Ms = 10, Mr = 200, so the bargain will bring more space for customers. The other related parameters are listed below: for convenience, the number of customers are set to 1000, the candidate hotels are 200 located in Chongqing City in China, data obtained from Ctrip from the year 2016. The distance of the hotel is obtained from Baidu Map using the direct distance around the train station (train station is in coordinate (0,0)). Level of the hotel is directly using the star level appraised by the hotel association. In Fig. 4, we present the result of price down of the bargain versus the average gain (utility function). In this figure, the x-axis means the normalized reduction of price, and the y-axis means the average gain of the customer through bargain, which is the utility function in the paper. We establish three groups of results in this figure.
123
S. Liu et al.
Fig. 4 Price reduction versus average gain
Every group contains theoretical result and Monte Carlo result. First, the red curve means the result using our proposed method with big data, the pink dotted curve represents the proposed method without big data, that means the value of policy Li(t) is fixed as constant, the blue dotted curve means the method of random game, which means the policy of customer and hotel are all random. From the result, we can infer that: 1.
2.
3.
Proposed game with big data perform the best, because the two-stage game allows the customer to have another chance to choose, and the hotel will also choose to accept the price, the probability that the second round bargain will decrease the price continue to increase while the second round starts. A maximum gain of 2 will be obtained through the proposed method; If big data information is deleted, the customer could not change the policy dynamically according to the historical data, only two bargain is introduced. From the figure, we can see that when the reduction is about 13, the method has the same performance as random game, and when the reduction increase, the performance is lower than random game, this is because the game is a noncooperative one, so the customer could not get enough information about whether the price is acceptable through two bargain, the probability that the customer reject a proper price will increase, which lead to the performance lost at the right side of the result. The Monte Carlo simulation result matches the data from theoretical analysis.
123
Game theoretic approach of a novel decision policy for…
Fig. 5 Average price reduction versus failure rate
In Fig. 5, we simulate the result of average price reduction versus failure rate. The x-axis indicates the average price reduction of the bargain, that means the normalized reduced price, the y-axis means the failure rate of the bargain, the rate is defined as the probability that the customer reject to stay at the hotel after two bargains. There are four results listed in this figure: our proposed method with big data (the dark curve with star), our proposed method without big data (red curve with rhombus), random game with big data (blue curve with circle) and random bargain without big data (pink curve with pentacle). From the result, we can infer that: 1.
2.
3.
With the increase of the price reduction, the failure rate increase, that means the hotel do not wish to reduce the price as the income will decrease, for different policies, the failure may be well controlled while the price reduction could keep stable; Our proposed method with big data perform the best, that is because the big data help decide the proper bound of price, so the failure rate could be well controlled, compared to the method without big data, we can see that there is about 5 times loss, so historical data is necessary; Random bargain receives the worst performance, for the control of rejection is rare, and some useful chances are wasted.
123
S. Liu et al.
5 Conclusion In this work, we discuss the problem that how a customer could choose a proper hotel to stay and receive the highest cost performance. A non-cooperative game between customer and hotels is introduced in this work. We first analyze the game model between customer and hotels, then we model the game with our proposed two-stage bargain game. Through the two-stage proposed policy, the customer could significantly reduce the final check-in price and the failure rate will not increase compared to random game. Acknowledgements This research was supported by the National Social Science Foundation of China (No.14AGL023). The work of Su Hu was jointly supported by the MOST Program of International S&T Cooperation (Grant No. 2016YFE0123200), National Natural Science Foundation of China (Grant No. 61471100/61101090/61571082), Science and Technology on Electronic Information Control Laboratory (Grant No. 162105003) and Fundamental Research Funds for the Central Universities (Grant No. ZYGX2015J012/ZYGX2014Z005). We would like to thank all the reviewers for their kind suggestions to this work. Authors’ contributions Shasha Liu and Bingjia Shao contribute the idea and the main algorithm of this paper, Yuan Gao and Su Hu perform the simulation of this manuscript. Weigui Zhou and Yi Li help improve the idea and writing of this work.
References 1. Adolph, M. (2014). Big data, its enablers and standards. PIK—Praxis der Information sverarbeitung und Kommunikation, 37(3), 197–204. 2. Asay, M. (2013). Q&A. Is open source sustainable? Technology innovation. Management Review, 3(1), 46–49. 3. Banerjee, S., Viswanathan, V., Raman, K., & Ying, H. (2013). Assessing prime-time for geotargeting with mobile big data. Journal of Marketing Analytics, 1(3), 174–183. 4. Barton, D., & Court, D. (2012). Making advanced analytics work for you. Harvard Business Review, 90(10), 78–83. 5. Bennett, M. (2013). The financial industry business ontology: Best practice for big data. Journal of Banking Regulation, 14(3/4), 255–268. 6. Bertot, J. C., Gorham, U., Jaeger, P. T., Sarin, L. C., & Choi, H. (2014). Big data, open government and e-government: Issues, policies and recommendations. Information Polity: The International Journal of Government & Democracy in the Information Age, 19(1), 5–16. 7. Bharadwaj, A., El Sawy, O. E., Pavlou, P. A., & Venkatraman, N. (2013). Digital business strategy: Toward a next generation of insights. MIS Quarterly, 37(2), 471–482. 8. Bone, S. A., Fombelle, P. W., Ray, K. R., & Lemon, K. N. (2014). How customer participation in B2B peer-to-peer problem-solving communities influences the need for traditional customer service. Journal of Service Research, 18(1), 23–38. 9. Chen, H., Chiang, R. H. L., & Storey, V. C. (2012). Business intelligence and analytics: from big data to big impact. MIS Quarterly, 36(4), 1165–1188. 10. Chow-White, P. A., & Green, J. G. (2013). Data mining difference in the age of big data: Communication and the social shaping of genome technologies from 1998 to 2007. International Journal of Communication, 7, 28. 11. Demirkan, H., & Delen, D. (2013). Leveraging the capabilities of service-oriented decision support systems: Putting analytics and big data in cloud. Decision Support Systems, 55(1), 412–421. 12. Dobre, C., & Xhafa, F. (2014). Parallel programming paradigms and frameworks in big data era. International Journal of Parallel Programming, 42(5), 710–738. 13. Fanning, K., & Grant, R. (2013). Big data: implications for financial managers. Journal of Corporate Accounting & Finance, 24(5), 23–30.
123
Game theoretic approach of a novel decision policy for… 14. Gabel, T. J., & Tokarski, C. (2014). Big data and organization design. Journal of Organization Design, 3(1), 37–45. 15. Galbraith, J. R. (2014). Organization design challenges resulting from big data. Journal of Organization Design, 3(1), 2–13. 16. George, G., Haas, M. R., & Pentland, A. (2014). Big data and management. Academy of Management Journal, 57(2), 321–326. 17. Reips, U. D., & Garaizar, P. (2011). Mining twitter: A source for psychological wisdom of the crowds. Behavior Research Methods, 43(3), 635–642. 18. McAfee, A., Brynjolfsson, E., Davenport, T. H., Patil, D. J., & Barton, D. (2012). Big data: The management revolution. Harvard Business Review, 90(10), 61–67. 19. Fan, W., & Bifet, A. (2013). Mining big data: Current status, and forecast to the future. ACM SIGKDD Explorations Newsletter, 14(2), 1–5. 20. Schroeck, M., Smart, R., Romero-Morales, D., & Tufano, P. (2012). Analytics: The real-world use of big data: How innovative enterprises extract value from uncertain data. Cambridge: IBM Institute for Business Value. 21. Vossen, G. (2014). Big data as the new enabler in business and other intelligence. Journal of Computer Science, 1(1), 3–14. 22. Geethakumari, G., & Srivatsava, A. (2012). Big data analysis for implementation of enterprise data security. IRACST-International Journal of Computer Science, Information Technology and Security (IJCSITS), 2(4), 742–746. 23. Rajaraman, A., & Ullman, J. D. (2012). Mining of massive datasets. Cambridge: Cambridge University Press. 24. Begoli, E., Horey, J. (2012) Design principles for effective knowledge discovery from big data, software architecture (WICSA) and European conference on software architecture (ECSA) 2012 (pp. 215–218). 25. Nardo, M., Petracco Giudici, M., & Naltsidis, M. (2015). Walking down wall street with a tablet: A survey of stock market predictions using the web. Journal of Economic Surveys. doi:10.1111/joes. 12102.
Shasha Liu received the B.S. and M.S. degree from Southwestern University of Finance and Economics, Chengdu, China in 2009 and 2012. Now she is a Ph.D. student at Chongqing University, Chongqing. Her research interests are Tourism Marketing, Tourism E-commerce, Big data in Tourism. Bingjia Shao is professor and Chair of Department of Marketing in the School of Economics and Business Administration at Chongqing University, China. He received his M.S. and Ph.D. degree in Southwest University of China in 1994 and 1997. His areas of interest are Electronic Commerce, Customer Relationship Management, and Internet Marketing. He has published in journals such as Information & Management, Electronic Markets, and Journal of Global Information Technology Management. Yuan Gao (S’09-M’15) received the B.S. degree from PLA Information Engineering University, China in 2008 and M.S. and Ph.D. degree from Dept. of Electronic Engineering, Tsinghua University in 2011 and 2014 and is currently an assistant professor in Tsinghua University and China Defense Science and Technology Information Center, China. His research interests are wireless communication system, satellite communication system, network control theory and big data. Professor Gao is a member of IEEE and ACM, he is an associate editor for several international journals, including IEEE Access, Eurasip JWCN and Sensors, etc. He is also a guest editor of several special issues. Professor Gao also serves as guest reviewer and TPC member of several journals and international conferences, including IEEE JSAC, Trans. on Wireless Communication, Trans.on Communication, IEEE Communication Letter, ICC, WCNC, and so on, he has published more than 40 academic papers in peer-reviewed international journals and conferences. Su Hu received his M.S. and Ph.D. degree in the National Key Lab on Communications from University of Electronic Science and Technology of China (UESTC) in 2007 and 2010, respectively. Currently, he is now an Associate Professor at the UESTC. From Feb. 2011 to Aug. 2012, He was a Research Fellow in
123
S. Liu et al.
School of Electrical and Electronic Engineering, Nanyang Technological University, Singapore. His research interests include sequence design with good correlation properties and physical layer design for wireless communication systems, such as filterbank multicarrier systems, cognitive radio networks. His homepage is at http://staff.uestc.edu.cn/husu/. Yi Li (M’15) received the B.S. degree from University of Electronic Science and Technology of China, Chengdu, China in 2009 and double degree in Economics from Peking University in 2014. She received her Ph.D. degree from Department of Electronic Engineering, Tsinghua University in 2014. During the year 2012, she was a visiting scholar of Columbia University.Yi is also a member of IEEE and CCF. Now she is currently working with The High School Affiliated to Renmin University of China, Beijing. As a mathematics and computer science teacher, she also focus on guiding students mathematical modeling and inquiry study in electronic engineering. Her research interests are wireless communication system, E-learning, Big data in Education, and STEAM Education. Weigui Zhou received the B.S. degree from Wuhan University, Wuhan, China in 2007 and M.S. degree in National University of Defense Technology in 2009. Now he is currently working with the Xichang Satellite Launch Center of China, Xichang. His research interests are wireless communication system, Big data in Computer Network.
123