(Memo) Requirement of an academic "Survey Paper"

Survey Paper is distinct from other type of academic papers such as something called Research Paper, Proposal etc. Then, since I haven't been able to find an effective set of "direction" type of document which explains what are the required criteria for an academic Survey Paper (if any), I've just collected a couple of references from internet and picked up some points discussed. I'm just listing those for my own purpose but also hoping it might help some others. Only in this blog entry, Survey Paper can be also called just Survey for brevity purpose.

I hope this blog entry itself could be a good "Survey" over the topic ;) (which could be NOT since this is just a random memo).

---

From http://drzaius.ics.uci.edu/wiki/index.php/Writing_survey_papers (bold letters are by author, indented are original)
New research results not needed; instead, relationship b/w research perspective needed

It is not intended merely to demonstrate that you have collected and read a lot of papers and books. Rather, it is intended to show that you can contribute to the research in that area. Generally, it will do this by giving some novel insight into the research area not through new research results (that comes later) but instead by demonstrating something of the relationships between research perspectives at work in the domain.

Survey is not a pure excerption from exisiting materials

what you are trying to do is to give the reader of the paper some insight into the domain. This insight will not simply be “extracted from” the literature, but rather arise from how you put the survey together.

Difference b/w Topic Proposal

This raises a complicated issue that frustrates many people, which is the relationship between a survey paper, your own research, and your future topic. Certainly, the survey paper is not a topic proposal. A topic proposal is a detailed description of a research project, intended not only to describe and motivate a set of research questions but also to set out precisely how they will be answered.

Survey should reflect your future research

Your survey will not contain this level of detail, by any means. However, it should, implicitly or explicitly, address future work. Its orientation towards this work is three-fold. First, it should delineate an area of study; second, it should demonstrate that interesting work remains to be done in this area; and third, it should show that you have some insight into this area that can suggests you will be able to find a dissertation-worthy topic. (snip) So, although the survey is not the topic proposal, some discussion of your future research is not excluded.
(snip)
while your survey should not be about your research, it can certainly explain how it is related to, or inspired, grounded, or shaped by your research to date.

Scoping is another hard part

choosing an appropriate scope – being able to identify and articulate a research area – is part of the challenge of writing a survey paper in the first place, (snip) You need to be able to say not just “my topic is X” but also “my topic is not Y or Z; they are related in such-and-such a way, but I am focused on X”

#pages be around 50, #reference should better be b/w 80 - 120

survey paper should be judged in terms of its intellectual coherence rather than “by the numbers.” (snip) In my own experience, a good survey paper tends to be around 50 pages of text, with between 80 and 120 references. When you have fewer than 50 references, I worry that you are not taking a broad enough view and not dealing a large enough sample of the literature to really be able to identify themes and trends. When you have over 200 references, I begin to wonder whether you have scoped your area of interest adequately enough to be able to say something concrete.

From "サーベイの仕方・論文の読み方" (In Japanese, "The way to survey, how to read academic papers", all translation by author of this blog)
This is not about a survey as a Survey Paper, but as a survey for the basis of one's own research.

- サーベイの流れ (Flow of survey)
1. 調査分野を決める (decide domain)
2. 資料を集める (collect materials)
-- セッション単位で抜き出すと良い (gathering per session is a good approach)
-- 広く浅くが大事 (breadth rather than depth)
3. 資料を読み，整理する (look through them, sort out)
4. サーベイ資料の形にする (sort out as survey)
5. 研究に反映する (reflect to your own research)
6. 1に戻る (Back to 1)

- サーベイ資料のまとめ方 (Key of sorting out survey)
分野の現状をまとめることが大事 (summarize the current situaition of the domain)
何が研究に役立つのかを明確に (clarify what can be the feedback to your research)
-- 現状，何が未解決か．問題か．(what's the problem unsolved)
-- 何ができていて，何ができていないか．(what's done / undone)
-- 社会的な要求は何か？ (requirement from society / industry)
わかりやすく，表にまとめる (Use table layout)

---

What's "Yakanuki Chicken" ? Might have meant "Yakiniku" - grilled meat @ Wholefoods Market, Dallas, TX.