Gene sequence similarity due to shared ancestry after a duplication event, i.e. paralogy, complicates the assessment of genetic variation, as sequences originating from paralogs can be difficult to distinguish. These confounded sequences are often removed prior to further analyses, leaving the underlying loci uncharacterized. Chum salmon have only partially re-diploidized subsequent to a whole-genome duplication and have loci with both di- and tetrasomic inheritance patterns. We present a maximum likelihood-based method to resolve confounded paralogous loci by observing the segregation of alleles in haploid offspring and demonstrate its effectiveness by constructing two linkage maps for chum salmon (<i>Oncorhynchus keta</i>): with and without these newly resolved loci. We find that the resolved paralogous loci are not randomly scattered across the genome, but are clustered in the sub-telomeric regions of 14 linkage groups, suggesting a significant fraction of the chum salmon genome may be missed by the exclusion of paralogous loci. Transposable elements have been proposed as drivers of genome evolution we find a reduced fraction of transposable element annotations among the resolved loci, suggesting their important role in the re-diploidization process by driving differentiation between homeologous chromosomes.