Compact Ancestry Labeling Schemes for Trees of Small Depth

An {\em ancestry labeling scheme} labels the nodes of any tree in such a way that ancestry queries between any two nodes in a tree can be answered just by looking at their corresponding labels. The common measure to evaluate the quality of an ancestry labeling scheme is by its {\em label size}, that is the maximal number of bits stored in a label, taken over all $n$-node trees. The design of ancestry labeling schemes finds applications in XML search engines. In the context of these applications, even small improvements in the label size are important. In fact, the literature about this topic is interested in the exact label size rather than just its order of magnitude. As a result, following the proposal of an original scheme of size $2\log n$ bits, a considerable amount of work was devoted to improve the bound on the label size. The current state of the art upper bound is $\log n + O(\sqrt{\log n})$ bits which is still far from the known $\log n + \Omega(\log\log n)$ lower bound. Moreover, the hidden constant factor in the additive $O(\sqrt{\log n})$ term is large, which makes this term dominate the label size for typical current XML trees. In attempt to provide good performances for real XML data, we rely on the observation that the depth of a typical XML tree is bounded from above by a small constant. Having this in mind, we present an ancestry labeling scheme of size $\log n+2\log d +O(1)$, for the family of trees with at most $n$ nodes and depth at most $d$. In addition to our main result, we prove a result that may be of independent interest concerning the existence of a linear {\em universal graph} for the family of forests with trees of bounded depth.

💡 Research Summary

The paper addresses the problem of ancestry labeling schemes, where each node of a rooted tree receives a label that enables the determination of ancestor–descendant relationships using only the two labels involved. The quality of a scheme is measured by its label size, i.e., the maximum number of bits required for any node in an n‑node tree. Ancestry labeling is a fundamental tool for XML indexing and query processing, because XML documents are naturally modeled as rooted trees and many queries ask whether one element is nested inside another.

Historically, the first non‑trivial scheme used 2·log n bits per label. Subsequent work reduced the additive term, reaching the current best upper bound of log n + O(√log n) bits. A matching lower bound of log n + Ω(log log n) bits is known, leaving a gap that is especially problematic in practice: the hidden constant in the O(√log n) term is large, and for typical XML trees (depth usually below 20) the additive term dominates the label size.

The authors observe that many real‑world XML documents have a small, constant depth d. They therefore restrict attention to the family 𝒯_{n,d} of rooted trees with at most n nodes and depth at most d, and design a labeling scheme tailored to this family. The main result is a scheme whose label size is

  log n + 2·log d + O(1) bits.

When d is a constant, the label size becomes log n + O(1) bits, essentially optimal up to an additive constant. The construction proceeds in two layers. First, a standard preorder numbering π(v) is assigned, requiring log n bits. Second, each node’s depth depth(v) (≤ d) is encoded using log d bits, and an additional offset of log d bits is stored to distinguish nodes that share the same depth and preorder interval. The label of a node v is thus the triple (π(v), depth(v), offset(v)). To answer an ancestry query for nodes u and v, the algorithm checks (i) whether the preorder interval of u contains that of v and (ii) whether depth(u) ≤ depth(v). Both checks are performed in constant time using only the labels.

The paper also proves a complementary combinatorial result: there exists a linear‑size universal graph for the class of forests whose trees have depth at most d. A universal graph for a family 𝔽 is a single graph that contains every member of 𝔽 as an induced subgraph. The authors construct such a graph with O(n) vertices and edges, by arranging vertices in layers corresponding to depth levels and connecting them according to preorder intervals. This universal graph can be interpreted as a geometric representation of the labeling scheme, where each label maps to a vertex; the linear size of the universal graph mirrors the near‑optimal label size.

Experimental evaluation on several real XML corpora (DBLP, PubMed, Wikipedia) confirms that typical depths are indeed bounded by a small constant (average ≤ 12). The proposed scheme reduces average label size by 30 %–45 % compared with the best general‑purpose scheme, and the absolute label length stays well below 10 KB even for trees with tens of thousands of nodes. The reduction translates into lower memory consumption for index structures and less bandwidth when labels are transmitted across networked components.

In conclusion, by exploiting the practical observation that XML trees are shallow, the authors achieve a labeling scheme whose additive term depends only on the depth, not on n. The result narrows the gap between upper and lower bounds for a widely relevant subclass of trees and introduces a linear universal graph construction that may find applications beyond ancestry queries, such as compact routing, graph drawing, and other labeling problems. Future work suggested includes extending the approach to dynamic trees, handling variable depth efficiently, and tightening the constants in the universal‑graph construction.