How difficult are decision problems based on natural data, such as pattern recognition? To answer this question, decision problems are characterized by introducing four measures defined on a Boolean function f of N variables: the implementation cost C(f), the randomness R(f), the deterministic entropy H(f), and the complexity K(f). The highlights and main results are roughly as follows. 1) C(f)...

We present an alignment-based approach to semi-supervised relation extraction task including more than two arguments. We concentrate on improving not only the precision of the extracted result, but also on the coverage of the method. Our relation extraction method is based on an alignment-based pattern matching approach which provides more flexibility of the method. In addition, we extract all ...

This paper proposes a pixel-based texture classifier that integrates multiple texture feature extraction methods in order to identify the regions of an input image that belong to a given set of texture patterns. Experimental results with textured images of outdoor scenes show that the proposed technique yields lower classification errors than widely recognized texture classifiers based on speci...

We present a method for learning wrappers for multi-slot extraction from semi-structured documents. The presented method learns how to construct automatically wrappers from positive examples, consisting of text tuples occurring in the document. These wrappers (T-wrappers) are based on a feature structure unification based pattern language for information extraction. The presented technique is a...

Cottrell, Munro & Zipser's (1987) proposal that their image compression network might be used to automatically extract image features for pattern recognition is tested by training a network to compress 64 face images spanning 11 subjects, and 13 non-face images. Features extracted in this manner (the outputs of the hidden units) are given as input to a one layer network trained to distinguish f...

We describe a general approach to the task of information extraction from free text and propose methods for learning syntax patterns automatically from annotated corpora. We study the application of our approach to the extraction of protein-protein interactions from scientific texts. Based on this evaluation, we find that learning patterns outperforms techniques based on handcrafted patterns.

The main focus of this work is to investigate robust ways for generating summaries from summary representations without recurring to simple sentence extraction and aiming at more human-like summaries. This is motivated by empirical evidence from TAC 2009 data showing that human summaries contain on average more and shorter sentences than the system summaries. We report encouraging preliminary r...

We propose a new design method, called discriminative feature extraction (DFE) for practical modular pattern recognizers. A key concept of DFE is the design of an overall recognizer in a manner consistent with recognition error minimization. The utility of the method is demonstrated in a Japanese vowel recognition task.

For biomedical information extraction, most systems use syntactic patterns on verbs (anchor verbs) and their arguments. Anchor verbs can be selected by focusing on their arguments. We propose to use predicate-argument structures (PASs), which are outputs of a full parser, to obtain verbs and their arguments. In this paper, we evaluated PAS method by comparing it to a method using part of speech...

Feature extraction is one of the major components in traditional pattern recognition. There are many methods for extracting the features, either structural approach or global approach. In this paper, we present integrated formulation of Zernike Moments and United Moment Invariant for extracting the character images accordingly. The extraction values are validated by measuring the Inter-class an...