The possibility of social simulation and big data mining may be feasible
November 16, 2022 10:07 Source: "China Social Sciences", November 16, 2022, Issue 2531, Issue 2531

Computing Social Science is considered to be developed along two parallel methods of ways。One is based on actress -based modeling (Agent Based Modeling,ABM) Social simulation path mainly,It mainly explores the emergence of macro -complex social phenomena (Emerge) mechanism and its non -expected evolution; the other is machine learning、Complex network analysis、Big data analysis path mainly based on text analysis,Mainly based on the increasing number of electronic traces in the past 20 years as the analysis data,By developing new methods, analyze the association and law behind social phenomena。Although the two paths have been developing in recent years in the development of research tools、Talent training、Knowledge production and other aspects have achieved rapid development,But I also encountered their respective development bottlenecks。The former was ridiculed as a "toy model" because of lack of sufficient data for model calibration and verification;,Ignoring the theoretical interpretation of the phenomenon and being criticized as "exquisite mediocre"。The former limited the expansion of research issues due to the lack of external data,The latter restricts the depth of its research due to the inadequate explanation of social theory。Facing the increasingly complex social phenomenon brought by digital transformation,The fusion potential of the two methods path is imperative,There are multiple possibilities,There are also several challenges。

 Bet365 lotto review

There are two main explanations about the separation of the development path of the development of social Bet365 app download sciences。A explanation is that the two appear differently。Social simulation has gone through system dynamics、Ear -cell automatic machine to three main development stages, including active -based modeling (ABM),The earliest can be traced back to the 1940s; and big data analysis mainly appeared in the last 20 years,Especially Lazele and others have developed the "Declaration of Calculation Social Sciences" in "Nature"。Another explanation is that there are large differences in the research groups of two methods and the issues they are concerned about。Social Simulation Pay attention to the emergence mechanism of macro phenomena,Focus on the complexity of the "micro -macro" connection,It has attracted the attention of a large number of sociologists early。and big data mining pay more attention to using the newly emerged mass trace data、Administrative Big Data and other resources conduct research on human behavior,The main goal of discovering the operating mode of macro social phenomena and precise predictions for it。Due to the visibility of big data resources、Method reserves and scene applications and other reasons,This field has long been led by a large number of computing scientists and physicists,Of course, as computing social sciences in economics、Sociology、The rise of traditional social sciences such as psychology has been changed to a certain extent。

Social simulation based on ABM pays attention to the macro phenomenon of long -term interaction of individuals and the environment.,By establishing a "artificial world" that can represent the real society,Set the interactive rules of the individual,Macro social phenomenon that emerges based on this rule -based interaction,such as social isolation、Cultural differentiation、Group intelligence, etc.。The biggest challenge bet365 best casino games for this method path is how to confirm how the interactive rules are the only rules for the emergence of macro phenomena。Different from nature,Social phenomena often exist in multiple mechanisms and multiple reasons,So,The simulation of social phenomena has increased the difficulty of this challenge。Researchers need to find the most critical and explanatory mechanism in theory and logic to expand the analysis。At the same time,Those who need to set the "artificial world" settings,The rationality of the attributes and interactive rules through data description,and also need to compare the simulation results with the real world development results to illustrate the rationality of the model。So,The demand for the "high -dimensional reality" model of the field of experience in this field is getting higher and higher,Data obtained high cost and research design complexity limits the extension of such models,At present, in addition to planning in urban planning、Infectious disease communication and other fields develop faster,There are still no large -scale applications in other fields。

At the same time,Big Data Analysis Path Although in full swing at the moment,Similarly facing insufficient explanatory dilemma。On the one hand,This path has been marked with commercial applications earlier,Researchers attach great importance to predictability,Even some people still think that "related is enough,No causal explanation "。On the other hand,It is that the big data of the trace is not a regular data,The value information contained in is even thinner,Methods using traditional assumptions to carry out interpretation research difficulties。With the in -depth research of traditional social science issues with big data analysis,This path is increasingly valued for explanation。bet365 live casino games For example,successively proposed a variety of ways such as "combination of big data and small data", "exploratory and verification comprehensive research", "large -scale experiments based on the Internet"。But,The emergence of big data not only indicates the innovation of the method,At the same time, it also means the transformation of human society,connection method of people、Connecting scale and social differentiation methods are undergoing significant changes。Our future will be a digital society with more complexity.,Complex social phenomena and non -continuous evolution are increasing daily。Understanding of these social phenomena,More and more needed to explain the "vertical cause and effect" specifically for complexity。

 Fusion path and their challenges

Facing the increasingly complex social phenomenon,The fusion potential of two methods is imperative。Social simulation that lacks external data calibration and big data analysis that lacks the interpretation of "vertical cause and effect" is in mutual fusion,Can make up for their weaknesses。The macro social model found and its associations found by rigorous big data analysis can provide data support for the model of ABM,At the same time, it also indicates the direction of discovering the new research problem。and ABM provides possible explanations for the social model and its associations found in big data analysis,Discover the emergence mechanism behind it,Even experimental research for possible intervention methods。Actually,"Handbook of Sociology Science" published in 2022,Several analysis sociologists have made a wonderful summary of the integration cases of this field。Overall,The method of fusion of the two mainly includes the following three types。

First,Big data analysis results as a basis for ABM modeling parameter Bet365 lotto review check as a single indicator。ABM's model setting is more flexible,The data set by different parameters can be diversified,Therefore, the results of big data analysis can enter the ABM model in different ways。such as,Analysis of the Big Data Registration Big Data can be used as the basis for the setting of the basic attribute settings in ABM,The analysis results of social network big data can be used as the basis for the setting of network parameter settings among the active network parameters in ABM。This is a relatively shallow method fusion,Big data analysis has assisted the role of completely auxiliary,The challenge is whether the big data analysis can provide unbiased、Precise、New indicators that cannot be provided by other survey data。

Second,Big Data Analysis and Social Simulation played a role in undertaking,To deepen the understanding of the problem together。A typical case comes from the social robot research carried out by Marijn A. Keijzer and Michael Mas。Existing big data analysis results show,Social robots play a contradictory role in information dissemination: the information of social robots spread is very wide,But at the same time, it can only affect a few people。Why can it only affect the social robot of a few people,Can the information spread widely? For this paradox,The author establishes an ABM model analysis based on the classic communication model,Those robots that quickly attract and change a large number of people around the people,In fact, a circle of communication between homogeneous views is established around it,also expanded the distance between the circle and the surrounding,thus prevented the long -range spread of information。In this case,The degree of integration of the two methods is obviously deeper,Big data analysis results provide research problems for ABM modeling,Between the bet365 live casino games two is a kind of undertaking relationship。Of course,This research can further verify whether the simulation results are consistent with the results of the real world through big data analysis。but but,Its challenge is that the problems raised by big data analysis require a mature、widely accepted ABM model to simulate。In fact,This kind of model is rare,This undoubtedly hindered the expansion of related issues。

third,Based on big data analysis, establish a "high -dimensional reality" ABM model。This type of research is what we are most looking forward to,Of course the challenge is also the largest。One of the challenges is the availability of the data,The ideal big data we think should be all -inclusive full data,In fact, even if there is such data at present,There are also available ethics problems。The second challenge is that big data is natural data,It may not be suitable for direct research。From this,Some cases appear in Internet experiment research。such as,Some researchers establish a music ranking website,The initial downloads of the songs in several experimental groups are ranked in different rankings,while the control group has no song download ranking。The "Matthew Effect" of the music market was tested through the above experiments,The popular song may not only be related to the quality of the song,It is also related to whether it is welcomed by others for its starting stage,is whether it is related to the "Matthew effect" mechanism。On the basis of this experiment,Arnout Van de Rijt establishes an experienced ABM model through experimental data,Further discovery through analog analysis,From the long term,Social impact process has a self -correction mechanism,Make the quality of the music consistent with its download ranking。The data of the research bet365 live casino games is completely based on big data analysis,The two analysis methods have reached a complete fusion。

(This article is the new path of the "new path to calculate the research and judgment mechanism of the ethnic group relationship in the era of social sciences" in Yunnan Province "Xing Dianying Cai" supporting the planning of young talents)

(Author Unit: School of Ethnology and Sociology, Yunnan University)

Editor in charge: Cui Cen
QR code icon 2.jpg
Key recommendation
The latest article
Graphics
bet365 live casino games

Friendship link:

Website filing number: Jinggong.com An Bei 11010502030146 Ministry of Industry and Information Technology:

All rights reserved by China Social Sciences Magazine shall not be reprinted and used without permission

General Editor Email: zzszbj@126.com This website contact information: 010-85886809 Address: Building 11-12, Building 1, Building 1, No. 15, Guanghua Road, Chaoyang District, Beijing: 100026