Title:Fully-Online Suffix Tree and Directed Acyclic Word Graph Construction for Multiple Texts

Abstract:We consider the construction of the suffix tree and the directed acyclic word graph (DAWG) indexing data structures for a collection of texts $\mathcal{T}$, where a new symbol may be appended to any text in $\mathcal{T} = \{T_1, \ldots, T_K\}$, at any time. This fully-online scenario, that arises when dynamically indexing multi-sensor data, is a natural generalization of the long solved semi-online problem, where texts $T_1, \ldots, T_{k-1}$ are permanently fixed before the next text $T_k$ is processed. We present fully-online algorithms that constructs the suffix tree and the DAWG for $\mathcal{T}$ in $O(N \log \sigma)$ time and $O(N)$ space, where $N$ is the total lengths of the strings in $\mathcal{T}$ and $\sigma$ is their alphabet size. The standard explicit representation of the suffix tree and the DAWG edges must be relaxed in the fully-online scenario, since too many updates might be required, and instead, we provide access to the frequently updated suffix tree leaf edge labels and the DAWG re-directable edges via auxiliary data structures, in $O(\log \sigma)$ time.