Computing Social Science is the product of the era of big data,Currently calculating social sciences is more understood as an emerging field for research on social science issues with calculation methods。February 2009,15 scholars led by David Lazer published a viewpoint article entitled "Calculating Social Science" on Science,marks the birth of this field。After 10 years,These scholars published an article on the Science Policy Forum again,Reflection and calculation of the shortcomings and problems in the field of social science research,Set some suggestions at the same time。As a brand new research field,Self -reflection by "10 Years Looking back",Several problems faced in the field of research,It is necessary for the future development of the subject。This article tries to calculate the relevant empirical analysis of social sciences over the past decade by reviewing the past ten years,Analysis of several disputes in the field of computing social science research methods,To present some challenges facing this field in this field。
Bet365 lotto review
Social sciences in the general sense,Still calculating social sciences,Its quantitative research is based on data。Look at the name alone,"Big Data" seems to emphasize that it has a larger amount/scale than traditional quantified data,But in the nature of the two、The relationship between the relationship with research is different。On the one hand,The nature of the two data is determined to be different。In general social sciences,Data is index value data; under the context of big data,The type of data and the complexity of the structure is much higher,both include text、Image、Video、Audio and other non -structural information,also includes space position、Highly complex information such as complex sequences。On the other hand,From the perspective of the relationship with research,Traditional quantitative data is used for experiments for specific research purposes、Questionnaire survey and other methods with planned observation results。In other words,The production of data Bet365 lotto review itself is part of research。But,Data mining of data production and social scientific research in the era of big data becomes two relatively independent processes,Research and data production relationships have been changed or even inverted -shift from production data to digging data。
Therefore,Compared to traditional small data,Big data does not mean more data,It also means that there is a way to produce、Form、dimension、Structure、Analysis method、The relationship between scientific research is completely different。From the perspective of methodology,Social science research based on big data,In fact, it represents a new social science research paradigm。
Dialectical view of objectivity
Some scholars think,"Naturally Occurring" is a basic attribute of social science big data。Whether it is the digital trace in the Internet、Information data in the Internet of Things,Still the email of the actor、Chat records, etc., are all considered to be naturally occurred in the state of unintelligo.,Reflected the real state of the actor。Compared with it,The data collected by traditional research methods will be obviously affected by the investigator intervention effect or self -report/social expectation deviation effect。Except for Internet of Things data mining,The above views are also accepted in some empirical research on social media data。This group of scholars will from Facebook、Twitter and other social networking sites collected by social networking sites are deemed to be "traces" (traces "or" symboms ",The website itself is considered to be a neutral service provider。This position pre -assumes that the relationship between the data and the user is self -explanatory,It is believed that the behavior of individuals (and groups) can be predicted through researching data。In the field of existing big data research,The prediction function is placed in an absolute priority position,Prediction through political and social phenomena to solve some society、Political and business issues。Relatively speaking,The function of the explanation is placed in a relatively secondary position,Even being ignored。
However,such as J. Dijck (2014) pointed out that the position implies such a belief,that is, quantitative is objective,It is possible bet365 Play online games to track everyone's behavior through online data,(yuan) Data can be analyzed and processed into a "raw material" for predictive algorithms about human future behavior。But,From the author's view,This concept cannot be based on the theory and methodology level。At the level of ontology,The objectiveness of the data claimed by this concept is not existing。For example,Facebook and Twitter constantly debugging,Friendship、Popularity, etc. convert to a certain algorithm,At the same time, this algorithm is called the "social" value concept。Website buttons such as "Praise" and "Hot Topics" may be considered as natural online social activities,But the algorithm that constitutes these buttons is carefully prepared to guide people to click the response。At the methodology level,Analysis and interpretation of data "raw materials",It will inevitably focus on a certain problem,and use a certain analysis tool。Therefore,To make the data mode meaningful,Need to criticize the geological question: Who do we have for the benefit、Why to find massive metadata。
I think,It can be divided into two places to look at the objectivity of big data。On the one hand,From the perspective of the production process,Compared to small data,Big data is a huge amount of information generated in the digitalization process of social life,Researchers have not intervened,Therefore, it is objective。On the other hand,Due to different types of big data generated channels、Production logic is different,Researchers can incorporate data production logic into consideration during the research process,Analysis of the impact of data production form on the conclusion of the research,You can also use the data cleaning method and algorithm to clean up the noise in the data。For example,As far as social network big data is concerned,There may be "Water Army"、Chat robots and other factors play an important role in the data production process,Then make the data full of false information, etc.,At this time, the researchers need to remove the information that cannot reflect the true public opinion or user characteristics through the algorithm。
Representative needs to be accumulated in the data
Not just objectivity,bet365 best casino games Representative (or deviation) is also an important issue in the debate of the nature of big data。In the book "Big Data Age",Merr-SCH (VIKTOR Mayer-SCH? 觟 NBERGER), etc. (2012), "More is not a random sample,but the overall data "is listed as the first content of big data thinking。They think,The arrival of the era of big data not only means an increase in data scale,Under certain circumstances, you can even collect all information about a phenomenon。So,Is the "overall" under the big data paradigm completely "all" in the real sense? if not,What is the representativeness of this "overall"?
For the above -mentioned "overall" theory view,E. Hargittai (2015) raised strong doubts。In his opinion,If the "overall" assumes the establishment,That means that it can be obtained through big data analysis、General conclusion,Even if the data used for analysis is limited to specific social networking sites or service items。But,No matter how large these websites or projects have, the scale,The research on these users will naturally miss or exclude non -user groups; and,User preferences of different websites、Features and experience of Internet technology (such as frequency and proficiency) will further weaken the representativeness of the data。
Some empirical research also shows the suspiciousness of the "overall" view,such as Brenner, etc. in 2013, a research report entitled "72%of adult netizens is social networking websites"。can be seen from the title,Number of netizens is not equal to the total population,And the number of users on social networking sites does not even equal to the total number of netizens。So say,Analysis of the expansion of only social network big data is obviously not a "overall" analysis。This can be seen,In the sense of "science",Big data may not be able to avoid more research errors than small data; in specific research,Careful derivation of the conclusion based on its sampling framework。
Is the analysis results based on big data? Or,Is the conclusion based on big data analysis unbiased? This question can be seen from two aspects。On the one hand,Different channels and data generated by platforms may be different from the overall extent。For example,In terms of social media big data,The number of users on some platforms may be bet365 Play online games closer to the overall,And the number of users on some platforms is just a part of society members。A possible research path is,Compare these data analysis results from different channels,or gradually approach the overall truth,Then test the stability and reliability of the analysis conclusion。On the other hand,From the perspective of long -term trends,Digital transformation of social life,The user scale of various platforms is also growing,Users gradually approach all users,At the same time, these platforms play a greater and more role in social life,So,The laws and characteristics of the social system described based on these big data are even more important。
Calculation and theoretical promotion
According to Anderson (2008),Because big data can "speak by yourself",Social Science is about to enter a "era of theoretical end"。This view has been questioned by many scholars once it was put forward。For example,"Big Data Age", although claiming to end causality analysis,in order to "let the data speak",But it is also acknowledged that the end of causal relationship is not equal to the end of theory,"The era of big data is definitely not an era of the theoretical demise; on the contrary,Theory runs through all aspects of big data analysis "。Theory is digging in data、Data analysis、Data understanding these three links all plays key characters。Data cannot "speak",and only after the analysis of theoretical guidance can we "speak"。Even the data mining stage,It also depends on the use of statistical tools and data models,and the establishment of the model、Selection of parameters, etc., are inseparable from theoretical guidance; the data excavated depends on theoretical speculation。But,Current big data research,Especially domestic big data research,quite superstitious "let the data speak" and ignore or even despise the theory。
Computer scientist Jim Gray (Jim Gray) proposed in 2007 that big data is the view of the "fourth research paradigm"。According to this point of view,There are experiments in human history、Theoretical deduction、Computer simulation three scientific discovery paradigms,In the era of big data, the fourth paradigm of the "data-inTENSIVE Scientific Discovery" is here。Although this bet365 live casino games view highlights the driving role of big data in the scientific exploration process,But does not deny the guiding meaning of the theory。The essence of the fourth paradigm is not to completely replace the experiments in the first three paradigms with big data、theory and simulation,Instead, unifying them with data on the new basis。The "Gray Law" in the fourth paradigm is the embodiment of the theory to play a leading role。
Actually,The relationship between calculation and theoretical is not just as simple as guidance and guidance,Calculation can also promote theoretical verification and development。Specifically,Computing Social Sciences provides the "Third Road" that is different from experimental laws and sampling survey laws for the test of conclusions and social theory in the past。Not only that,Expressing over time,Social science big data with strong timeliness will gradually become more time sequences、Super high -dimensional "vertical data"。From the perspective of scientific discovery,These "vertical data" contain huge theoretical value。
Promote the integration and development of multiple research methods
With the arrival of the era of big data,Is the traditional experimental method and sampling survey method outdated? Is the original understanding path "theoretical hypothesis -model -inspection" routine?、Sample survey method、Methods for content analysis by manually coding and calculating paradigm under the big data framework (such as machine learning、neural network、Natural Language Treatment) Combine,and emphasize the complementarity between these different research methods。But the empirical research in many big data paradigms tends to think,Traditional research methods、Research logic and understanding path is basically outdated。With massive data and powerful complex algorithms,Researchers can make precise predictions in many research fields without theoretical premise。Another researcher thinks,The theoretical driving of the traditional analysis method should be combined with data driving,Two -way interaction,Organic integration Promote computing social science research。But some scholars think,With the further development of artificial intelligence,Calculating paradigm is very likely bet365 best casino games to produce a revolution in the field of social sciences,Paradigm based on small data -based analysis paradigms may not return to the world。
To sum up,As an emerging comprehensive research field,Calculating the development of social science not only faces the problem of whether big data can effectively portray the operation of social systems and conduct effective analysis of social facts,At the same time in the research method,Also existing relationships with traditional research methods or competition or complement each other。Controversial issues in these research methods,Researchers can only answer in the practice of social sciences。In big data、Technology such as cloud computing and artificial intelligence is used in the research process of social sciences,Methodism of calculating social sciences will gradually become mature。
(Author: Institute of Finance and Economics of Hubei Institute of Economics and Economics; School of Social College of Wuhan University)
Friendship link: The official website of the Chinese Academy of Social Sciences |
Website filing number: Jinggong.com Anmi 11010502030146 Ministry of Industry and Information Technology: Beijing ICP No. 11013869
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: 11-12, Building 1, Building 1, No. 15, Guanghua Road, Chaoyang District, Beijing: 100026
>