direkt zum Inhalt springen

direkt zum Hauptnavigationsmenü

Sie sind hier

TU Berlin

Page Content

General Guidelines And Organizational Matters For Seminars

This one does always apply: If you have questions you can always ask your advisor!

Team work

In scientific (and industrial) research, it is common to work in a project together with others as a team. Constructively reading other peoples papers and the discussion about unclear aspects helps everyone involved.

The advisors of the seminar have only very limited possibilities for such a cooperation because they usually have a different view on a topic than a student. For this reason the participants of the seminar are divided into small groups (with as related topics as possible). The groups should work together as described below:

  • communication via e-mail or meetings, if questions occur,
  • mutual proof reading and commenting the papers after a first version had been submitted,
  • discussions about the papers with the supervisor as a moderator,
  • coordinating the content to avoid overlapps,
  • rehearsal talks (dry runs) with the other group members, if requested

After the first versions of the papers have been submitted, each participant has some time to proof read the papers of some of the other participants. That also includes:

  • thorough reading
  • annotating comprehension questions
  • marking mistakes (both, in content and grammatically)
  • evaluation of the content (comprehensibility, level of particularization, Ginger factor, etc.)
  • evaluation of the form (structure, formatting, bibliography, etc.)
  • overall impression

The scheduled time exposure for the proof reading adds up to 10 hours that should be split equally to the papers. Please, bring your proof read versions to the meetings.


For a graded certificate we expect a paper about 10 pages. (This default should neither be exceeded nor shortened a lot.)

Making the paper

  • Consider what parts of the literature you read should be described in your paper. Talk about it with your supervisor. If details are left out you should refer to corresponding literature.
  • Prepare a template for the paper. (The structure is the same as for scientific papers:)

    • title, name, date, location (which seminar, which university, which research group/chair, etc.),
    • abstract (summarizes briefly the whole work),
    • introducing chapter (with a motivation for the paper and an overview about the rest of the work),
    • if applicable a chapter with background information that is necessary for the unterstanding of the topic.
    • main part (can stretch to several chapters),
    • closing chapter with a summary,
    • if applicable an appendix – only rarely necessary!,
    • bibliography.

    We would appreciate you to use our LaTeX template (corresponding ps file) for your paper. (Requires only elementary LaTeX knowledge.)

  • Use chapter headings and different font sizes to distinguish between different levels of headings.
  • Think about, which parts of the literature should be covered in which chapter, and insert it into the template.
  • Do not stick too much on the given order and syntax of the literature. In particular, it is not enough just to read and translate the article. Here is transfer benefit in demand!

    1. You should do some background research. Don't just talk about the main paper in your paper but also about the whole subject area, that the main paper defines.
    2. To say it clear: A bibliography with only two items is a clear evidence that you did not deal sufficiently with your topic.
    3. Less important details can (should!) be left out. The basic ideas are more important than the implementation details.
    4. Things that are difficult to explain—in particular those you did not clearly understand yourself—should be explained in more detail.
    5. If necessary insert relevant background information into the famous "section 2" (background).
    6. If you insert formulas into your paper, make sure that these formulars are explained in detail, so that the reader can understand them. In most cases it is not necessary to insert formulas at all. Better explain the basic idea in words.

  • Fill the gaps in your template bit by bit.
  • Use consistent terms throughout your whole paper—even if you refer to different papers with different terminologies.
  • Variables in mathematical formulas are italicized, functions and operators are not.
  • Any literature used for your paper should be listed in your bibliography. On the other hand, any literature listed in your bibliography should also be cited in your paper.
  • All tables and images have to have captions: tables have captions above the table, images below.
    Each table and each image have to be referenced in the text. If you don't reference it you can also leave it out.
  • If you reach the fourth level of your outline ("section"), then your outline is broken.
  • Usually, two sections should not follow directly each other; e.g., you need at least one introducing sentence between "3.1 – Problems" and "3.1.1 – errors in protocols".
  • Try to meet the others' standard of knowledge: do not presuppose that they read what you read, but also do not begin at the very beginning if the basics are already covered by other participants of the seminar.


The scheduled time for each talk is 30 minutes plus 15 minutes for discussions.

Benefit for the student who gives the talk

  • You learn techniques to work non-trivial material and prepare a presentation autonomously,
  • to deal with criticism,
  • constant adaption to the audience's behaviour during the talk.

Benefit for the audience

  • Understanding the talk without / instead of reading the related literature,
  • practice constructive criticism. (It is not a sign of politeness but more a sign of misunderstood solidarity or own shyness, resp., if you do not play an active part in a seminar.)

Making the presentation / slides

  • Consider which parts of your paper should be introduced in your talk and talk about it with your supervisor.
  • Prepare a template for the slides:

    • Title slide: Topic, name, date, location (which seminar, which university, which research group/chair, etc.),
    • short introduction, motivation, outline of the talk (a table of contents, so to say),
    • main part,
    • conclusion/summary,
    • bibliography (if necessary already on the title slide)

  • Fill gaps bit by bit.
  • The Introduction should give an intuition about the topic, but it should not show any details, yet.
  • The content presented in your slides should only be content that you deal with in your talk.
  • How many slides? – There is no general answer to this question. It depends on your personal likes and dislikes if you want to go fast or slowly through your slides. Empirically, you need about 10–20 slides for a 30 minutes talk.
  • Graphics / pictures are helpful to clarify many statements. However, cliparts that do not directly deal with your talk act to be artificial to ridiculous.
  • Use consistent terms throughout your whole talk, even if you have several original papers with different terminology. (Then you should preferably use the same terms as used in previous talks.)
  • It can be meaningful to insert a short transition slide between the larger parts of the talk (introduction, main topic 1, main topic 2, or the like) for a few seconds. This shows the audience that a new part begins.
  • The words on the slides need to be written / printed in large letters (standard font: 24 pt), preferably in color. If you use a beamer you should use a sans-serif-font (i.e. without "little feet"), e.g., Lucida Sans, or Helvetica.
  • The usage of color is meaingful to emphasize important words and statements, or to establish a connection. However, please be consistent with each color.
  • No entire sentences on slides – people are not able to read a lot while listening to a talk. By no means just copy parts of your paper (e.g., theorems) to your slides!
  • In mathematical formulas, variables are put italic, names of functions, and operators are not.
  • Try to meet the others' standard of knowledge: Do not presume what you read yourself, but also do not begin with the very beginning, if someone else already held a talk about the basics. (Use the preparatory meetings to impress the supervisors :-))
  • The conclusion should complete your talk and should summarize its sense. If necessary, it can be meaningful to hint at advanced topics covered in your paper.

A talk has to be structured differently from your paper

  • Introductory part in the beginning with motivation and a lead to the topic.
  • Repeat what you are doing every once in a while – you should not expect the audience to exactly remember everything you said. And the listener certainly cannot "scroll back".
  • Summary at the end of the talk.
  • First drafts of the slides should be discussed in time before the talk, so you have enough time to change the global didactical concept (structure of the talk, etc.).

Time Of the Talk

  • Prepare a 30-minute talk with slide. Because of questions and discussions the talk will be extended automatically.
  • Even if you have very much material: no audition will be able to concentrate much longer, anyway.
  • It is quite an art to cut your material down, and you should practice it. On conferences you often have only 20 minutes, in business life even less—and people mean it for real.
  • We will quit your talk time after 45 min. mercilessly.
  • We know from experience that 10–15 slides are meaningful: most people do not manage to explain more slides within 30–45 minutes. This certainly depends on how much, or how less, resp., you write on each slide.
  • To estimate your time a rehearsal talk for friends would be best. Or, you can silently speak the very sentances you would say in your talk sitting before your computer. For a rough calculation you can find out how long you talk on avarage about a few representative example slides.

Translations English↔German

If you want to translate english terms to german you should keep in mind the following:

  • For many terms a good and established german term does not exist, but the english term has become very common instead.
    Example: "Routing".
  • Nouns are written upper-case in German. – For this reason English nouns are also written upper-case if used in a German text, even if they would be written lower-case in an English text.
    Examples: "Router", "Probe", "Traceroute", "Client".
  • For some (other) terms a unique german word does exist and should usually be used.
    Examples: "Netzwerktopologie", "Paketklassifikation", "Datenbank".
  • There are no composite nouns in English, but they are common in German. You can adopt fixed English terms which consist exclusively of English words as several words without hyphens to your German text (Example: "Route Flap Damping"); but you do not have to—often, hyphens make it more clear—or you use quotation marks.
    Example: "Route Flap Damping", "Route-Flap-Damping", or ""Route Flap Damping"".
  • If you combine several English terms to a new object you may not use spaces but you should keep it together with hyphens.
    Example: "Peer-to-Peer-Client" (Only English words; hyphens instead of spaces, though.)
  • It is also very common to use a German-English mixture as a term—this results from literal translations of English terms. Note, in this case the words are never written seperated. Instead they are either written in one word or they are seperated with dashes—a mixed term is always considered as a "German" term when writing in German. Therefore, all German rules are also applied to mixed terms. And in German, you don't write "Leber Wurst", "Spiegel Ei", "Braun Kohle" or "Kraft fahr Zeug"

    • Right: "Routingalgorithmus", "Route-Flap-Analyse", "BGP-Nachricht", "Peer-to-Peer-Client", "IP-Adresse", "Datenbankserver"
    • Wrong: "Routing Algorithmus", "Route Flap Analyse", "BGP Nachricht", "Peer-to-Peer Client", "IP Adresse," Datenbank Server

    Also have a look at www.deppen leer zeichen.de and the corresponding explanation at Wikipedia

  • The plural of English terms that is usually used as singular is constructed by German (!) rules. English plural rules are orthographically wrong!
    Example: eine History, "mehrere Historys" (und nicht: "Histories"—unless you use throughout the whole text solely the plural, then you can introduce the plural term as fixed term).


If you are unsure ask your supervisor.

Zusatzinformationen / Extras

Quick Access:

Schnellnavigation zur Seite über Nummerneingabe

Auxiliary Functions