1: \begin{abstract} 

2: A new relaxed variant of interior point method for low-rank semidefinite 

3: programming problems is proposed in this paper. The method is a step 

4: outside of the usual interior point framework. In anticipation 

5: to converging to a low-rank primal solution, a special nearly low-rank 

6: form of all primal iterates is imposed. To accommodate such a (restrictive) 

7: structure, the first order optimality conditions have to be relaxed 

8: and are therefore approximated by solving an auxiliary least-squares 

9: problem. The relaxed interior point framework opens numerous possibilities 

10: how primal and dual approximated Newton directions can be computed. 

11: In particular, it admits the application of both the first- and the 

12: second-order methods in this context. The convergence of the method 

13: is established. 

14: A prototype implementation is discussed and encouraging preliminary 

15: computational results are reported for solving the SDP-reformulation 

16: of matrix-completion problems.

17: \end{abstract}