I am a beginner in preparing beamer presentations. I don't know the advanced options available in it that make it more appealing and more sensible. Here is my code for the beamer presentation. I need suggestions on how to improve it further:


\setbeamercolor{section in head/foot}{bg=Brown}
\setbeamercolor{author in head/foot}{bg=Brown}
\setbeamercolor{date in head/foot}{fg=Brown}

\title[Semantic Web ranking]{Ranking of Web Documents using Semantic Similarity}
\author{Subham Soni}
  Submitted to\\
  Dr. (Mrs.) K. Saruladha\\ B.Tech., M.Tech., Ph.D.\endgraf\medskip
  Dept.\ of Computer Science and Engineering \endgraf
  Pondicherry Engineering College}}
\item To relate documents that have similar meaning, semantic similarity approach is used by matching the keywords which are extracted
\item Here, all the relevant relations between the keywords describing user's context and then calculating the fraction of these relations on each web page to determine their relevance to user's query
\textbf{Proposal:} A Ranking algorithm for the \textit{semantic web \mbox{documents}} by finding the semantic similarity between the \mbox{documents} and the user's query
\item A huge of information is stored in the WWW in the form of \emph{interlinked} web pages
\item Traditional Search Engines use \textit{lexical matching} and \textit{link analysis} to produce the results
\item To overcome the limitation of lexical matching, it is necessary to determine the \textbf{relationship} between the keywords
\item Other methods such as Ontology based matching, Semantic Analysis, etc. exist but here a novel ranking model is explored which provides the result-set according the user query and semantics of the document
\section{Related Work}
\begin{frame}{Related Work}
\item Existing Ranking models:
\item Boolean Model
\item Statistical Model
\item Hyperlink Based Model
\item Conceptual Model
\item Ontology based similarity
\section{Proposed Ranking Model for Semantic Web Documents}
\begin{frame}{Proposed Ranking Model for Semantic Web Documents}
\item Determine the keywords of the document using syntactic analysis to make vector space model of the documents
\item A domain specific dictionary is prepared
\item The words in the dictionary are assigned weights based on their relevance to the domain using \emph{fuzzy set approach}
\item Then, the mapping of each of the words in the vector space model stored in the document repository is done with the domain specific dictionary
\item The document similarity w.r.t the user query is also
found by extracting the concepts and the relationship that
exists between the concepts present in a document and the
mapping of the same is then done by the ontology processor
\item Finally the maximum value gives the overall relevance of the document with respect to the query
\item Consider four documents w.r.t education domain
\item The dictionary based weights are as shown:
\item The ontology based weights are as shown:
\item The documents considered are :
\item[D1:]Education is a lifelong process. A person learns through
his experience. It goes on forever from his birth to death
without any break or barrier.
\item[D2:]Education of man does not begin at school but begins at
birth. It ends not when he graduates from university but ends
at his death. Hence, Education is a lifelong process.
\item[D3:]Education is not only academics but social also. It is
important in one's person life.
\item[D4:] In a person life everyone needs to be educated and social.
Everyone learns through experiences gained in one's life.
\section{Future Work}
\begin{frame}{Future Work}
  \item To design more meaningful and exhaustive ranking strategy by using the semantic analysis of web pages and by deeply statistical analysis relevance of documents
  \item To make the approach scalable to the semantic web

The images I have used are the following:
enter image description here
enter image description here
enter image description here
enter image description here
The tables could have been prepared in TeX but due to the lack of time, I had to copy from the document and paste it.

  • Well the presentation looks fine. Perhaps you only need to insert overlays for the items of the lists and things like that. And of course the tables inside if you have time. – Aradnix Aug 21 '14 at 2:19

"Make it more appealing and more sensible" is a very subjective task.

I definitely prefer another themes than CambridgeUS with these colors, but I have no idea of the theme & color which is good for you. Without trial and error, you can make a quick comparison at the Beamer Theme Matrix page or in the Beamer theme gallery. For some custom themes see Where to find custom beamer themes and the templates of ShareLaTeX or Bala's Website.

Of course, you can also take a lot of ideas from the questions of this site (see tag beamer) and the documentation (run texdoc beamer).

Do not use \date{} to print address or any long text in the title frame. Note that is also partly printed in the right bottom corner of remaining frames upon the navigation icons!

One dirty trick to avoid this problem is use \date[]{interminable bla bla bla bla...} but for another stuff that the date use another commands as \subtitle{}, \institute{}, \titlegraphic{} or \logo{}. For example, change your very long \date by:


\titlegraphic{Submitted to\\
  Dr. (Mrs.) K. Saruladha\\ B.Tech., M.Tech., Ph.D.}

\institute[Pondicherry College]{
Dept. of Computer Science and Engineering \\
  Pondicherry Engineering College}


Note that like in \title, you can use the optional argument ([...]) in \author and \institute to control what is showed in the bottom bar. By the way, the top and bottom bars have no sense in the first slide, since all the information is already showed by \maketitle. Thus, instead of \maketitle try with \frame{\titlepage \thispagestyle{empty}}. Also, \useoutertheme{infolines} do not hurt, but is redundant with \usetheme{CambridgeUS}.

But without doubt, a major improvement (with or without LaTeX) will be remove all the vertical rules (and use only 3 horizontal rules) in the tables. Or may be use soft row colors instead of horizontal rules (but I think that row colors are better for highlight some data). Tip: use LaTeX with the booktabs package to use commands as \toprule instead of \hline. It worth spend time on it. For coloured rows you can use the colortbl package.

Another major problem: Too much text. Remember: People must follow your dissertation, but they are too busy trying to read the small text to pay attention. The text should not be a parallel version of your entire speech, and of course, do not read the text! (if you do not want to bore everyone).

If you really need include long lists with long texts, and intelligent use of overlays, alerts, boxes and fancy transitions of slides could help to stop the readers, so they can pay attention to you sometime, but note than even these effects could be more distracting that helpful.

Speaking of distractions: Really you need the navigation bar? Try with \beamertemplatenavigationsymbolsempty. Even more, try a minimalist theme like keynote

In summary, I think that instead of this ...


... with the proper use of preamble commands you must obtain at least that the number of the slide and the date are showed correctly (and of course, that the logo is not "Dep"):


But behind of the "technically correct", I think that is better change the first four lines of your code to ...

% \usetheme{CambridgeUS}
% \useoutertheme{infolines}

... and add these packages to the preamble ...


...so you can show the table in two clearer slides:

\def\grco{} do nothing in first slide
\only<2>{\def\grco{\rowcolor{green!40}}} % \grco put green color in second slide

\caption{Dictionary based weights of four documents.}
S No. & 
\parbox{.2\linewidth}{\centering Domain (Education)} & 
\parbox{.2\linewidth}{Weight assigned} & 
Synonyms \\
      1 &  Process    & .8 & $\cdots, \cdots, \cdots $\\ 
      2 &  Study      & 1 & $\cdots, \cdots, \cdots $\\%
\grco 3 & Learning   & 1 & $\cdots, \cdots, \cdots $\\ 
      10 & University       & .6 & $\cdots, \cdots, \cdots $\\ 

... the first with a nice but raw LaTeX table (without additional text: is your work explain how it is obtained...), and the second one highlighting the most important data (while you are stressing about their importance):


Remember: People must follow your dissertation, not the fireworks from your presentation. Images help. Short sentences and keywords help. The number of the slide may be help. A mini article do not help. You should not read/explain what they are already seeing. What they are seeing should explain what you're saying.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.