Empirical Study Of Microblogging Marketing Effectiveness Based On BP Neural Network

In March2006, Ivan Williams created a Twitter, microblogging marked the birth. It hasbecome a site worth nearly10billion U.S. dollars. August4of2009, the largest portal Sinalaunched the beta version of "Sina microblogging" website portal to offer micro-bloggingservice. Then, in a short span of more than a year, quickly set off a wave of domesticmicroblogging and officially entered the Chinese vision of mainstream population. Sinamicroblogging has become a leader in the field of domestic micro-Bo, now, it has more than200million registered users, tens of thousands of celebrities from all walks of life, a millionbusiness and institutional accounts. The First Year of Microblogging Market White Paper datashows that July2010, with the increasing of the number of users of the microblogging, thetotal number of Sina microblogging generated over90million microbloggings daily and theproduction is more than300million, successively, Tencent microblogging is also a goodmomentum of development. According to the authoritative institutions, which predict that by2013, the domestic microblogging market will be matured. Admittedly, microblogging isdeveloping rapidly by its unique advantages and charm characters, many domestic andforeign enterprises have been awared of the business opportunity, then the microbloggingmarketing is gradually increased, so that the the microblogging marketing practice activitiesare developed very quickly. In response to this phenomenon, the domestic and foreignscholars has began to study microblogging marketing.On the basis of summing up the results of the research of scholars at home and abroad,that is not difficult to find, foreign research is focused on the practice of micro-Bo in thevarious institutions, as well as in various fields of applied research, in favor of quantitativeresearch. But as for domestic research, the majority of studies still remain standing businesspoint of view and the microblogging marketing strategies exploratory study level, tend toqualitative research. So, whether the conclusions of the study abroad can be established in thecountry or not, whether the domestic microblogging marketing quantitative research can bestudyed, has become a hot topic.Read the relevant literature and found that the advantages and disadvantages ofmicroblogging marketing, both of two are seriously departured, the "profit" can lead to thebutterfly effect, to promote their products quickly, almost zero cost, it is an important reasonfor many businesses to settle microblogging. But there are many companies ignoring the "disadvantages" of microblogging marketing, the event of a crisis, and the speed ofinformation dissemination, beyond imagination, then public relationship has become a majorproblem. How much the impact of the microblogging marketing end to corporate marketing,has become a pressing problem.This study treated it as a research topic, through research of microblogging marketingtheory, connected the mode of the AESAR (Note (Awareness)-participation (Engagement)-attitude (the Sentiment)-action (Action)-reserved (Retention)) with the tripartite monitoringsoftware-the "Microblogging Billboard", theory with practice, analysis summarized theestablishment of the microblogging marketing effectiveness evaluation model, and proposedthe application of LM-BP (Levenberg Marquadt-Back Propagation) neural network model forthe implementation of the verification process, on this basis, then put forward microbloggingmarketing effectiveness evaluation index system, the calculation method and calculationformula. Selected microblogging account instance which has operational definition andcalculation, used the neural network toolbox of the Matlab mathematical modelingsoftwarethe to measure data validation and take prediction, and at the same time, used thegray forecasting software to make the long-term projections of data, used measured data andresults data, through the regression analysis, which demonstrated how much themicroblogging marketing has effect on their overall marketing.Compared with the research results, this study insights in the following ways:First, this study is the breakthrough of the shortcomings which the research is notstudyed deeply enough, acrossed from exploratory research to empirical research;Second, this study maked up the disadvantage of the domestic marketing microblogginglack of quantitative research, maked it more scientific;Third, it is the more in-depth study of the microblogging marketing effectiveness,combined the AESAR mode with third-party monitoring software, built the model, andproposed the application of the LM-BP neural network model to verify the implementationprocess;Fourth, the establishment of the microblogging marketing effectiveness evaluation indexsystem, put forward the corresponding calculation methods and formulas;Fifth, as for the selected instance, operational definition, estimation, application of theMatlab mathematical modeling software neural network toolbox and gray modeling softwarefor the predictive analysis, regression analysis made more scientific evidence, has draw theconclusion that microblogging marketing has no significant effect on their overall marketing effectiveness;Sixth, the application of the popular topology principle, neural network and grayprediction fuzzy information processing, to provide a reference, also become the direction ofnew research ideas;Seventh, the application of different mathematical modeling software, such as the Matlabmathematical modeling software and gray prediction software for model validation andprediction, the application of appropriate statistical analysis software, such as Eviewsstatistical analysis software, for measuring scientific data and making the results moreconvinced.There are three points of shortcomings in the course of the study: the amount of data isnot enough, the weights are independent assignment, take FMCG as a case study, for otherindustries,whether it can get the same conclusion or not, to be studied.