The basic function or mission of Research Gate is doubtless highly beneficent. Important is not only the possibility to offer own publications to a large number of potentially interested persons, but also the feed-back data showing what articles are red, downloaded and cited. However, a long-termed experience with these data indicates that something does not work correctly in Research gate.

Still about one year ago the number of readings or downloading of papers was high, with a considerable predominance of readers from USA and China. Well, it seamed to be logical; USA and China are huge countries, with an enormous number of researches. BUT, my topics are relatively narrowly limited to Central Europe and hardly can be interesting for people from remote areas. I did not understand it.

About one year ago the situation changed. Number of read or downloaded paper declined rapidly. This had a simple logic. Number of interested researches is not infinite and their interest became saturated. BUT at the same time the readers from USA and China almost disappeared. Well, an also logical consequence, at least in case of geographical limiting of my topics. But it also looks like an effect of geographical filtering. In addition, there was always a strikingly low representation of readers from of other scientific power – Russia.

BUT – now there appeared other question. Among an about 3-4-times lower number of read or downloaded papers, (absolutely among 10-15 papers), 1 or 2 thematically specialized paper, in some cases even published in languages that are understandable only to few readers, often highly dominated. One of such papers was frequently downloaded even 8-times. In each days some other of such papers. There I have only two more or less logical explanations – the counter of reads works improperly or the readers have difficulties with downloading and each unsuccessful attempt to download a paper is recorded.

Sometimes I get question if a paper is of me, in spite of fact that it already is in my profile. There are only two possible answers – yes and it will be added duplicate – or not and I will lie. A filter would be desirable.

The recording of citations is very good, sometimes it comes even much sooner than in other databases. But there is recorded a huge number of autocitations. A filter would be very desirable. The profile data would be more realistic and the list of cited papers easier to use.

Do you have a similar experience?