Semiquandles and flat virtual knots (Q600667)

Virtual knots were introduced by \textit{L. H. Kauffman} as a combinatorial generalization of classical knots [Eur. J. Comb. 20, No.~7, 663--690 (1999; Zbl 0938.57006)]. A flat virtual knot diagram is derived from a virtual knot diagram by forgetting the over and under crossing information at all real crossings. By doing the same to the generalized Reidmeister moves of virtual knot theory we obtain moves for flat virtual knot diagrams. Flat virtual knots are then defined to be flat virtual knot diagrams modulo these moves. A flat virtual singular knot diagram is a flat virtual knot diagram where some real crossings are marked singular. Additional moves which involve singular crossings are defined and flat virtual singular knots are flat virtual singular knot diagrams modulo these moves. The number of singular crossings is invariant under these moves. The knot quandle was defined independently by \textit{D. Joyce} [J. Pure Appl. Algebra 23, 37--65 (1982; Zbl 0474.57003)] and \textit{S. V. Matveev} [Math. USSR, Sb. 47, 73--83 (1984; Zbl 0523.57006)]. It is an algebraic structure derived from a classical knot diagram where the generators correspond to arcs of the diagram and relations are given by the crossings. The axioms on the quandle are derived from the Reidmeister moves in such a way that the quandle is a knot invariant. The quandle was generalized to virtual knots by \textit{L. H. Kauffman} [loc. cit.]. Generalizations of this idea include the biquandle (see for example, \textit{L. H. Kauffman} and \textit{D. Radford} [in: Radford, David E. (ed.) et al., Diagrammatic morphisms and applications. AMS special session on diagrammatic morphisms in algebra, category theory, and topology, San Francisco, CA, USA, October 21--22, 2000. Providence, RI: American Mathematical Society (AMS). Contemp. Math. 318, 113--140 (2003; Zbl 1031.57009)]) and the virtual biquandle, cf. \textit{L. H. Kauffman} and \textit{V. O. Manturov} [Fundam. Math. 188, 103--146 (2005; Zbl 1088.57006)]. In this paper the authors introduce an algebraic structure called a semiquandle. A semiquandle is a set with two binary operations which satisfy certain axioms which correspond naturally to the moves on flat virtual knot diagrams. The fundamental semiquandle of a virtual link is an invariant which is derived from any diagram of the link. The generators are the edges between real crossings and each real crossing generates a relation. A singular semiquandle is a semiquandle with another two binary operations which satisfy certain axioms which are derived from moves involving singular crossings. The fundamental singular semiquandle of a flat singular virtual link is defined similarly to the fundamental semiquandle of a virtual link. It is an invariant of a flat singular virtual link. \textit{L. H. Kauffman} and \textit{V. O. Manturov} introduced the virtual biquandle in [loc. cit.]. It is a generalization of the biquandle obtained by adding a unary operation which is applied at virtual crossings. The paper under review defines virtual semiquandles and virtual singular semiquandles by adding the same operation to semiquandles and singular semiquandles and defining certain axioms. Any flat virtual knot has an associated invariant virtual semiquandle and any flat singular virtual knot has an associated virtual singular semiquandle. Invariants, called counting invariants, can be derived from any of these kinds of semiquandles by counting homomorphisms from the semiquandle to finite semiquandles. Examples are given in the paper. Kauffman first considered Vassiliev (finite type) invariants of virtual knots by extending the definition of Vassiliev invariants of classical knots. In [J. Knot Theory Ramifications 19, No.~4, 461--487 (2010; Zbl 1195.57030)], \textit{A. Henrich} defined two Vassiliev invariants for virtual knots of degree \(1\) called \(S\) and \(G\). In that paper Henrich proved that \(G\) is stronger than \(S\). An alternative proof of this fact, using counting invariants derived from the singular semiquandle, is given in the paper under review.