사회과학과 컴퓨터사이언스
웹으로 필요한 정보를 찾다가 University of California, Riverside 학교의 Dr. Hanneman의 웹페이지 (soc109)을 방문하게 되었습니다. Class 과제에 대한 페이지였는데 과제 중에 Relational Databases에 관한? 항목이 있길래, “이게 Computer db 에서 말하는 rdbm과 같은 뜻일까?”하며 읽어 보았습니다. 과연, 이에 대한 언급이었습니다. “사회학부에서 relational databases에 대해서, 그리고, 이런 데이터베이스를 짜보는 연습을 시킨다…” 대단한 듯 싶습니다. 아래는 해당과제의 전문입니다.
This assignment does not require actual hands-on computer work (other than to write-up your answer and submit it for grading). Rather, I would like you to think about a research problem that one of the Professors here is working on, and offer advice about how a relational data base program might assist her in her work.
The researcher with whom we are consulting is involved in a study of mergers between large American corporations. She is interested in collecting information about a large number of mergers that occurred in the 1980s, so that she can identify patterns of involvement by certain individuals, kinds of individuals, and organizations. The information that she needs to keep track of is quite complicated: there are a large number of mergers (several hundred); in each merger, two or more companies are involved — but the same company can be involved in one or more mergers over the period we are studying; we are also interested in the individual people involved in making deals: who are the members of the boards of directors of each firm, who is the CEO and CFO (chief executive officer, chief financial officer), and outsiders who play roles in making the deals happen (e.g. brokers and bankers). Again, individual people can be involved with more than one company, and more than one deal. Sometimes, the same individual can play different roles (our researcher has identified three “roles”: buyer, seller, and broker) in different deals (being a buyer in one, a seller in another, and a broker in a third).
Our researcher would like to be able to record all of this information efficiently (i.e. making as few repeated entries of information). She would like to be able to quickly and easily answer questions like: how many times was Fred Smith involved in mergers, what roles did he play? Of the deals that were made by company Z, how many of them had brokers who were not members of the board of directors?
Your job is this:
Design (you do not have to actually build one) a relational data base that will help our researcher efficiently store and retrieve the information she needs for her study. Your design should specify: what tables of data will be set up, and what kinds of information you will store in each table; it should specify how the tables will be linked, and what kind of relation (one to many, one to one, many to one) is involved in each link. And, your design should provide and example of 1) a data entry form 2) a query, and 3) a report that you might develop to help the researcher.
http://faculty.ucr.edu/~hanneman/soc109/homework.html#Relational%20data%20bases 에서 인용
사실 이와 같은 데이터 준비작업은 사회과학을 하는 사람도 꼭 필요하게 되었습니다. 웹이나 블로그 등의 콘텐츠 분석 등에서 구할 수 있는 데이터의 규모가 얼마나 방대해 질지 모르는 상황이니만큼, 효과적인 데이터베이스 구축이 필요할 수 있으니까요. 이를 위해서 rdbm을 이용하여 데이터 테이블을 효과적으로 분해해서 설계할 계획과, 이를 효율적으로 다시 불러올 수 있는 계획을 작성하라는 과제입니다 (실제 데이터베이스를 짜라는 과제는 아닙니다).
더욱이, Dr. Hanneman이 이것을 과제에 넣은 이유는 social network analysis 쪽에서 계속 연구를 하여 온 사람이기때문입니다. 사회과학도 이제 컴퓨터사이언스가 필요한 시대가 왔나 봅니다 왔습니다. 이제 사회과학도도 sql을 배워두면 좋을 듯 합니다.
