5. Conclusion

The development of alternative metrics is particularly recent, and they must be taken with great caution, particularly in the absence of any real validation of these metrics as impact indicators. Further research into alternative metrics is needed before comparisons can be validated, and before technical difficulties and limitations can be overcome. The development of these metrics will also necessarily depend on the uptake of social networking by scientists, which is still very marginal. However, the main challenges will be more social than technical, and will concern the need to better understand the different logics of audience, whether disciplinary or between scientific and non-scientific sources, as well as to clarify the definition of impact and the uses to which its measurement is put. Many players are already particularly vigilant with regard to the central issues of data quality...