research process - How should I proceed if the author of a paper doesn't reply to my query?

During my literature survey for a project, I came across papers that have very interesting results and have decent number of citations. Unfortunately, the sub-area I am exploring doesn’t have any standard datasets on which I can test my implementation of the algorithm presented in their paper. The authors have generated the dataset by applying certain actions on a subset of publicly available information, which are Flickr images in my case.

So I have contacted the authors, asking whether they can provide me with their code or the dataset that they used for their implementation so that I may use it to perform my experiments and possibly try to extend their approach. But almost a week has passed by, and I haven’t received any response from them. Does that mean that,

1. My method of asking was not correct? (In that case, what would be appropriate?)


2. They don’t have the code or dataset available? (The papers were published in 2004.)



3. They are not interested?

Would it be polite to remind them again?

Answer

The reason for no response may be anything from what Anthony Labarre mentions to what you list. I would wait at least a month before writing off an answer; I have received answers even later than that. It is possible they are working on improvements themselves as well. In short, if you do not get a reply, you can only try again and perhaps express your request differently.

You seem to have contacted all the authors. Sometimes the project leader may be too busy to pick up on questions and the younger team members may be more inclined to answer. They may, however, want to talk it over among themselves, and it is probably not a high priority.

So try again in a couple of weeks. If you do not get any response, then there is probably not much you can do. You probably need to think about what can do to improve the chances for a reply. The word "collaboration" is sometimes a good way to "sweeten the deal". Sometimes, just to give a perspective, I get requests for data that are of the sort, give me this or that, give me everything, and I often end up asking myself, why should I? having laboured to generate the data. I am not saying you must follow such an approach to be successful but considering why the other researchers would want to help you may also help towards solving the communication problem.