cuc College of Information and Communication thesis sharing session

1. Literature study

1. Paper search

• Find literature URL
  (1) http://libw.cuc.edu.cn/
  (2) https://sci-hub.shop/
  (3) https://www.researchgate.net/
  (4) http:/ /www.4243.net/

• Find literature APP:

• Find the library of papers
  (1) Google Scholar;
  (2) IEEE Digital Library (cuc has purchased this library, so you can download papers);
  (3) Arxiv: In addition to some published papers, there are many unpublished papers also uploaded Go to this website;
  (4) ACM Digital Library: ACM conference papers are in this library, but cuc did not purchase this library, you need to find a solution;
  (5) Web of Science: query SCI papers, but our school does not have Purchase this library;
  (6) SCI-hub: It can be used to download SCI papers, but the full text website link or DOI of the paper is required for search;
  (7) Some good quality blogs, public accounts, and Zhihu columns (SIGAI, intelligent unit, Intensive learning knowledge lecture hall).
  ps: The selection of documents is mainly based on the number of citations and publication time of the documents, so that higher quality/newer work can be screened out.

ps: The library is a database of documents, established by some institutions, and the library will include journals approved by these institutions. These stocks are in the form of URLs.
But if we download documents directly from the library website, it usually costs money. At this time, the school needs to buy the library, and then we download the articles in the library through the school website. But if there is no way to download it for free, you need a third-party URL (for example: sci-hub), through this sci-hub, you can download the journals included in sci (sci is a library) for free.

2. Time planning
   (1) Intensive reading (print→notes→notes): 1-3 articles/week
   (2) Extensive reading/browsing: 4-6 articles/week
   ps: Use fragmented time to read. Broaden the scope of knowledge, understand cutting-edge knowledge, and inspire inspiration.

3. Literature reading
   (1) Read Chinese first, then English.

(2) Read the abstract and conclusion first. While reading, pay attention to the author and the institution, the place where the paper is published, and select the papers you need.

(3) How to read non-review literature?
➀Read it the first time, quickly browse it, you can use the mind map to sort out the structure of the whole article;
➁Read the second time, pay attention to the method in the paper, the experimental design, pay attention to how the author leads the problem, how to design the experiment to verify your own Point of view, some good experimental design plans need to be recorded;
➂Read the third time, pay attention to the author’s English expression, and extract some good sentences;
➃For every paper you have read, be sure to use Two or three sentences can summarize the entire article, and use a document to classify and record the papers you see. This is very important! If you need to write a summary/research background in the subsequent topic or other times, you only need to extract it from this document.
➄Regardless of the stage, always check the latest literature.
➅After reading a lot of the paper, use various charts to summarize and summarize to deepen the impression!

(4) How to read review literature?
➀While reading, use mind maps to organize the framework of the entire paper.
ps: Some commonly used mind mapping tools: Mindmaster (http://www.edrawsoft.cn/mindmaster/, an online
mind mapping production site) Visio, Xmind, etc.

(5) How to find related papers on this topic?
➀Looking back: look for the references of review papers/papers that have been found, you can find some newer classic papers in the field for reading, and then look forward from the references of these papers, through this The method can be searched back along the timeline, which probably establishes the development context of the subject.
➁Search back: After finding a document that is more suitable for your research topic, you must pay attention to the article citing this paper, so that you can see what other people have done on the basis of this paper, and whether there is still You can continue to study some points that others have not paid attention to.

2. Thesis writing

ps: Communicate the direction of the thesis with the supervisor

1. Preliminary preparation
   (1) Develop experimental skills as soon as possible: For many directions that require experiments in science and engineering, we must practice more in the early stage and use laboratory equipment. You don't have to be eager to get a good result early. In my opinion, it is good to have a negative result at the beginning, and I am afraid that I am confused and do not pay attention to details. The purpose of this point is to improve one's experimental skills as soon as possible. If you can't control the details of the experiment of the equipment you use, even if you have a good idea later, it will consume a lot of energy to implement it.
   (2) In normal experiments, you must pay attention to conceiving your own thesis structure, design experiments with a purpose, have your own guesses before the experiment, verify your guesses through experiments, and analyze your own guesses based on the experimental results. Yes/No, so as to constantly revise the design of the experiment, so as to avoid doing a random experiment. When I finally wrote the paper, I found that the experiment did not have any logical relationship to connect it.
   (3) Detailed record of experimental data: In the usual experimental process, for future repeatability, a notebook must be prepared to record various experimental parameters. The results of multiple performances can be summarized in the form of a table. Conducive to horizontal/vertical comparison, the pictures generated by the experimental data during the experiment (if it is difficult to generate) must be saved with high resolution (greater than 300dpi).

ps: Accumulate English expressions in normal times: classification and summary, such as professional vocabulary in the field, expressions of diagrams, tables, formulas, etc.

2. Paper structure
   (1) The main tools for paper writing: Word/Latex (1. Installation environment; 2. Editing software: VScode, Texmaker, etc.; 3. Direct use of tex web page editing software, such as Overleaf (https://www.overleaf. com/) etc.) You can find paper templates from the Call for Paper on the conference/journal’s official website. Usually Word/Latex templates are given. It is recommended to use Latex to avoid consuming a lot of effort due to format adjustments, and the typeset papers are more beautiful and increase Probability of acceptance of large papers.

3. Points for Attention in Paper Writing
   (1) Table: The header is above the table, and the presentation is clear;
   (2) Pictures: record data in the usual experiment, remember to save high-resolution pictures when saving pictures, drawing tool: matplotlib library (using python language, Pay attention to line shape, color, etc.), PPT (very easy to use, system block diagram/network structure diagram, etc., saved as PDF or high-resolution picture), Excel (experimental results are summarized into bar graphs, line graphs, etc.), Visio
   (3) pictures And the table should be placed at the top of each page as much as possible;
   (4) The font that appears in the picture should be consistent with the font used in the paper (if the font used in the paper template is special, use times new roman for the font in the picture), and try to Ensure that the size of the font in the picture is about the same size as the font of the paper. If the font is not conspicuous enough when the picture is placed in the paper, it needs to be bold;
   (5) Focus on the abstract and the introduction part (this is the main part of the review during the review) , Pay attention to the logical relationship between each sentence, the research significance of the topic, the introduction of the problem, and the main contribution/innovation of your thesis must be clearly written! The logical relationship of the whole paper must be clear, and the sentence transition must be smooth.
   (6) The wording of scientific research articles should be modest and prudent. It can be said that the literature [XX] did not do any work / did not consider XX factors and so on. It cannot be said that there are errors in the research of [XX] literature. It is very likely that the author of the literature you commented on is you If you give inappropriate comments, your paper will never be turned back in the hands of "his (her)";
   (7) Omit important references (indicating that the current situation analysis is not comprehensive), give People have a feeling of incomplete understanding of related fields;
   (8) The introduction does not summarize the inheritance and innovative relationship between the research in this article and the existing research;
   (9) The issue of normative references in references;
   (10) Language issues (mainly grammar ), good articles, 40% language, 60% innovations, new conclusions, etc.;
   (11) The result analysis should be comprehensive and effective, and logically rigorous: Many people feel that the paper has good results, but they are ruthlessly rejected. A large part of the reason may be the lack of logic in the paper structure and result analysis, and the article is too confusing. It is difficult for reviewers to understand what the paper is trying to say. For example, in the previous narrative, XX simulation/analysis was done, but it was not found at all in the results; or in the introduction of the paper, XX model was established in order to solve the XX problem (question 1, problem 2). However, in the analysis of the following calculation examples, we only see that the model can solve problem 1 but not problem 2. At this time, if the inconsistency is inconsistent, we need to modify the introduction or modify the model to make it consistent. Therefore, conclusions and abstracts can be left at the end to avoid inconsistencies.
   (12) Figure: Journal papers, >600dpi, generally need to be redrawn (excellent data needs to be presented to reviewers with beautiful pictures), tools (MATLAB, Python, Origin, PPT, Visio, PS)
   (13) Very important, very important, very important for suitable journals! ! ! Selecting journals is comparable to choosing female (male) friends. The paper's direction must be basically consistent with the journal field before it can be submitted for review by the editor.
   (14) A good blog: http://blog.sina.com.cn/s/blog_51b45a680101czt7.html

4. Checking and repeated revisions after the paper is completed
   (1) Open the spelling check in Word, and find the simplest word spelling mistakes and other low-level errors;
   (2) Grammar check tools (such as Grammarly, etc.), check the grammar of the entire paper, making the full text smooth ;
   (3) to read, confirm each sentence no ambiguity, try not to use too long sentences and multiple clauses put together especially prone to ambiguity, is not conducive to reading;
   (4) figures, tables, formulas, whether the meeting /Journal requirements (whether it is clear and beautiful, and good-looking charts will add color to the paper);
   (5) Whether each variable used in the formula explains its physical meaning, whether the symbols used before and after the same variable are consistent, whether each a sentence
   (6) References whether normative, try not to choose unpublished papers, comparative algorithms in the experimental section typically need to include classical algorithms in the field and the latest algorithm;
   (7) academic integrity: experimental thesis true Reliable and repeatable! (Very important)

3. Paper publication

1. Search common websites for submitting conferences/journals
   (1) WikiCFP: http://www.wikicfp.com/cfp/
   (2) Call4Papers: http://www.call4papers.cn/C4PREST/index.html
   (3) Find conferences Is it supported by IEEE: https://conferences.ieee.org/conferences_events/
   (4) CCF recommended conference list: https://www.ccf.org.cn/xspj/gyml/

2. When searching for conferences, you should focus on
   (1) The topics of the paper submissions listed on the conference’s official website. Try to choose a conference that fits your thesis topic. If your thesis topic and these columns of the conference are not suitable, there will be a lot of Risk of rejection.
   (2) Call for paper: Check the important time points of paper submission and the conference's requirements for submitted papers, and prepare your own papers according to this requirement.
   (3) The location of the conference and whether the conference requires authors to attend the conference.
   (4) Whether conference papers can be indexed by EI.
   ps: Selected journal submission: Ask about the review time of the combined journals (the better the journal, the longer the review time), publication fees, etc.

3. Waiting for the paper submission results
   Rebuttal: Some conferences will have a Rebuttal link, this link is that the review will ask you some questions/suggestions/questions based on your paper, and may ask you to provide more comprehensive materials and data, or request additional For some supplementary experiments, you need to carefully consider how to respond, and try to convince the reviewer that your experiment is reasonable/innovative. After the Rebuttal session, wait for the final employment results.

4. Paper results
   (1) If the paper is not accepted: adjust your mentality, it is normal for the paper not to be accepted. The submission of the paper is regarded as a trial and error process. Through the submission, you can know the problems of the paper. According to the review comments Correct the paper and prepare for the next submission.
   (2) If the paper is accepted: you still need to modify the paper in accordance with the review comments, and prepare the final camera-ready version of the paper (that is, the paper for final publication) according to the publication requirements of the conference party. This link needs to be more careful Check carefully to confirm that there are no errors, because once the camera-ready version of the paper is submitted, there is no chance to modify it. If there are errors, it will be very troublesome to not correct them.

5. Participation
   There are two ways to display papers: poster/oral presentation
   (1) Poster: Make a poster, and post it at the venue when attending the meeting. The meeting usually has a special poster display session. Stand next to the poster in this session. If someone tells you If you are interested in work, you need to explain your work to them; the principle of preparing a poster: use more diagrams and less text. PPT can be used as the production tool, pay attention to the length and width and the horizontal/vertical version.
   https://guides.nyu.edu/c.php?g=276826&p=1846154 (poster production process and template)
   (2) Oral: oral report. When attending the meeting, give an oral report (general international meetings require English). The time is generally 10-15 minutes. After the report, there will be about 5 minutes for questioning.

6. Things to prepare before the meeting:
   (1) You need to prepare a PPT in advance, be concise and generous, and explain your research ideas clearly (ask yourself why, and sort out why you want to do this before you can convince the audience);
   (2) Be prepared Speech drafts, improvisation is not recommended;
   (3) Practice oral and listening English in advance, and try to avoid situations where the questioning session does not understand other people’s questions;
   (4) If it is a foreign conference, there will usually be a lot of cumbersome before participating Preparatory work, such as various school filing materials, visa application, etc. It is recommended to prepare one to two months in advance, be patient, and keep electronic versions of various bills/materials.

Four, time node:

1. Research One
   (1) Class:
   (2) Papers:
   ➀Read a lot of literature, conduct summary planning, build a preliminary knowledge framework, and start experiments as soon as possible,
   ➁consolidate knowledge through practice, and find the points that you will study in depth;
   ➂publish conference papers.

2. The second semester
   (1) Design and work:
   last semester: write a literature review, find research points for the final design, prepare for the opening
   of the semester , prepare for the data structure and algorithm, and prepare for the job interview; next semester: finish Set most of the content, and pay attention to the company's summer internship recruitment at the beginning of the semester (internship conversion is easier than autumn recruitment).
   (2) Papers:
   ➀Complete the simulation work of one's own subject and obtain valid data;
   ➁Writing and submitting journal papers (more time-consuming). Second research, evaluation of national scholarships→
   October➂Preparation for the topic (end of second research, December)

3. 
   Graduate
   school three (1) final design: last semester: autumn recruitment, job hunting (internet, banking, public institutions, etc., different directions require different preparations), at the end
   of the semester, the graduation thesis should be prepared (preparation for blind review); next semester: After finishing the design, continue to write the graduation thesis and prepare for the final defense.
   (2) Papers:
   ➀Complete the simulation work of one's own subject and obtain valid data;
   ➁Writing and submitting journal papers (more time-consuming). Second research, evaluation of national scholarships→
   October➂Preparation for the topic (end of second research, December)