Inference algorithms for pattern-based CRFs on sequence data
This work provides incremental improvements for researchers in sequence tagging and structured prediction, offering faster inference algorithms for pattern-based CRFs.
The paper tackles the problem of efficient inference for Conditional Random Fields (CRFs) with pattern-based potentials on sequences, presenting algorithms that improve complexities for partition function, marginals, and MAP computation, with complexities such as O(n L) compared to previous O(n L |D|).
We consider Conditional Random Fields (CRFs) with pattern-based potentials defined on a chain. In this model the energy of a string (labeling) $x_1...x_n$ is the sum of terms over intervals $[i,j]$ where each term is non-zero only if the substring $x_i...x_j$ equals a prespecified pattern $α$. Such CRFs can be naturally applied to many sequence tagging problems. We present efficient algorithms for the three standard inference tasks in a CRF, namely computing (i) the partition function, (ii) marginals, and (iii) computing the MAP. Their complexities are respectively $O(n L)$, $O(n L \ell_{max})$ and $O(n L \min\{|D|,\log (\ell_{max}+1)\})$ where $L$ is the combined length of input patterns, $\ell_{max}$ is the maximum length of a pattern, and $D$ is the input alphabet. This improves on the previous algorithms of (Ye et al., 2009) whose complexities are respectively $O(n L |D|)$, $O(n |Γ| L^2 \ell_{max}^2)$ and $O(n L |D|)$, where $|Γ|$ is the number of input patterns. In addition, we give an efficient algorithm for sampling. Finally, we consider the case of non-positive weights. (Komodakis & Paragios, 2009) gave an $O(n L)$ algorithm for computing the MAP. We present a modification that has the same worst-case complexity but can beat it in the best case.