The software and knowledge engineering literature de nes maintenance strategies for seven main types of knowledge: words; sentences; behavioural knowledge; and meta-knowledge. Meta-knowledge divides into problem solving methods; quality knowledge; x knowledge; social knowl-5 edge; and processing activities. There are ve main ways in which these seven knowledge types are processed: acquire; operationalise; fault; x; and preserve. We review systems that contribute to these 7 5 = 35 types of knowledge maintenance. FixK SocialK QualityK Processing Activity MaintenanceKB PSM ExecutionKB SentenceK WordK BehaviouralK X W Y Z Y isa X Z may use X (and visa versa) KEY W can be partOf X maintained maintenanceKB executionKB by a second focus of most of current KA Figure 1: Maintenance KBs are a special kind of execution KBs which access more types of knowledge. Much of modern KA is focused only on execution KBsand preserve. These ve processing strategies and seven knowledge types de ne the 35 points of a knowledge management options space (see Figure 2). As we 45 move from acquiring wordK (\K" for knowledge) to preserving socialK (top-left to bottom-right of Figure 2), we see progressively less research work. Note the empty points in Figure 2 (marked by a \-"): these gaps are due to the recursive maintenance problem. Due to the fragmented and diverse nature of the KM eld, this article does 50 not aim to cover every KM system. Rather, we will mention systems that illustrate portions of the maintenance strategy option space 1 . Many of the systems reviewed here contribute to multiple places within the maintenance strategy option space. Hence, their descriptions may be broken up into di erent sections. Note that this article is \the view from NSW University" and hence 55 will have certain biases. For example:Much of modern KA focues on acquiring and operationalising PSMs.