I am PhD student works in a computer science. I have given lots of presentations in the institute level and one or two at the conference level. I have been told by my fellow researchers that they did not get much from my presentations/talks. Even I have presented one paper in the conference also and got the same feedback. I spend more time on making slides as well practise at least two three times on the projector also. I think I am kind of person who is worst at presenting the work. Assuming this Is there any way through which I can deliver a good talk?
I'm also not great at presenting. One method I've found useful is to present to and get feedback from people who have around the same level of knowledge as your intended audience (i.e. people who are not too heavily involved in your project, but are in your research field). This means they will be able to tell you which parts you need to explain in more detail.
Also, look at other people's presentations to see what's the standard level of technical detail in your field - I've seen presentations where the focus should be on the scientific analysis and results, but instead too much time is spent on the minutia of how the analysis was performed.
The truth is you need more detailed feedback than "I didn't get much out of it". Why, and when, did you lose your audience? Did they not understand your problem statement? Did they not understand your approach? Was it a language problem? Were you speaking too fast, too softly, or too unclearly? Did you use too much jargon without explaining it? Was the talk too technical (or, much more rarely, not technical enough)? Was the talk too boring to pay attention to?
What you need is to practice your talk in front of a small audience of people who are willing to give you direct and detailed feedback, ideally slide-by-slide. Only once you know what the most important problems are you can work on resolving them.
If your problem is literally about conveying your idea (i.e., describing the technical contribution of the paper in an understandable manner) my experience is that people undervalue two concepts:
- Having a really good illustrative example. Ideas are hard to follow if they are presented in the abstract, with As and Bs and Ns. Have a good, plausible (if small) example, and show how your approach works for it.
- Having good visualizations (that are made specifically for the presentation, not just a figure from the paper). Powerpoint animations have a bad rep, but for some types of research (e.g., some algorithms) a couple of animations make it much easier to work out what's going on than a code listing or a couple of text bullets.
I find Simon Peyton Jones' talks immensely useful. They offer practical advice and tips, and are catered to a CS audience. Apart from that, as @Emma says, there isn't really much cleverness to it - just practice until you're at least a passable speaker. Solicit feedback from your peers and advisors, and don't get discouraged.