Evaluating retraining rules for semi-supervised learning in neural network based cursive word recognition

Training a system to recognize handwritten words is a task that requires a large amount of data with their correct transcription. However, the creation of such a training set, including the generation of the ground truth, is tedious and costly. One way of reducing the high cost of labeled training data acquisition is to exploit unlabeled data, which can be gathered easily. Making use of both labeled and unlabeled data is known as semi-supervised learning. One of the most general versions of semi-supervised learning is self-training, where a recognizer iteratively retrains itself on its own output on new, unlabeled data. In this paper we propose to apply semi-supervised learning, and in particular self-training, to the problem of cursive, handwritten word recognition. The special focus of the paper is on retraining rules that define what data are actually being used in the retraining phase. In a series of experiments it is shown that the performance of a neural network based recognizer can be significantly improved through the use of unlabeled data and self-training if appropriate retraining rules are applied.

Automatic Labeling of Software Components and their Evolution using Log-Likelihood Ratio of Word Frequencies in Source Code

As more and more open-source software components become available on the internet we need automatic ways to label and compare them. For example, a developer who searches for reusable software must be able to quickly gain an understanding of retrieved components. This understanding cannot be gained at the level of source code due to the semantic gap between source code and the domain model. In this paper we present a lexical approach that uses the log-likelihood ratios of word frequencies to automatically provide labels for software components. We present a prototype implementation of our labeling/comparison algorithm and provide examples of its application. In particular, we apply the approach to detect trends in the evolution of a software system.

Recognition memory across the lifespan: the impact of word frequency and study-test interval on estimates of familiarity and recollection

The goal of this study was to investigate recognition memory performance across the lifespan and to determine how estimates of recollection and familiarity contribute to performance. In each of three experiments, participants from five groups from 14 up to 85 years of age (children, young adults, middle-aged adults, young-old adults, and old-old adults) were presented with high- and low-frequency words in a study phase and were tested immediately afterwards and/or after a one day retention interval. The results showed that word frequency and retention interval affected recognition memory performance as well as estimates of recollection and familiarity. Across the lifespan, the trajectory of recognition memory followed an inverse u-shape function that was neither affected by word frequency nor by retention interval. The trajectory of estimates of recollection also followed an inverse u-shape function, and was especially pronounced for low-frequency words. In contrast, estimates of familiarity did not differ across the lifespan. The results indicate that age differences in recognition memory are mainly due to differences in processes related to recollection while the contribution of familiarity-based processes seems to be age-invariant.

Unconscious relational inference recruits the hippocampus

Relational inference denotes the capacity to encode, flexibly retrieve, and integrate multiple memories to combine past experiences to update knowledge and improve decision-making in new situations. Although relational inference is thought to depend on the hippocampus and consciousness, we now show in young, healthy men that it may occur outside consciousness but still recruits the hippocampus. In temporally distinct and unique subliminal episodes, we presented word pairs that either overlapped (“winter–red”, “red–computer”) or not. Effects of unconscious relational inference emerged in reaction times recorded during unconscious encoding and in the outcome of decisions made 1 min later at test, when participants judged the semantic relatedness of two supraliminal words. These words were either episodically related through a common word (“winter–computer” related through “red”) or unrelated. Hippocampal activity increased during the unconscious encoding of overlapping versus nonoverlapping word pairs and during the unconscious retrieval of episodically related versus unrelated words. Furthermore, hippocampal activity during unconscious encoding predicted the outcome of decisions made at test. Hence, unconscious inference may influence decision-making in new situations.

Low reliability of perceptual priming: consequences for the interpretation of functional dissociations between explicit and implicit memory

In this study, three experiments are presented that investigate the reliability of memory measures. In Experiment 1, the well-known dissociation between explicit (recall, recognition) and implicit memory (picture clarification) as a function of age in a sample of 335 persons aged between 65 and 95 was replicated. Test-retest reliability was significantly lower in implicit than in explicit measures. In Experiment 2, parallel-test reliabilities in a student sample confirmed the finding of Experiment 1. In Experiment 3, the reliability of cued recall and word stem completion was investigated. There were significant priming effects and a dissociation between explicit and implicit memory as a function of levels of processing. However, the reliability of implicit memory measures was again substantially lower than in explicit tests in all test conditions. As a consequence, differential reliabilities of direct and indirect memory tests should be considered as a possible determinant of dissociations between explicit and implicit memory as a function of experimental or quasi-experimental manipulations.

Unconscious word stem completion priming in a mirror-masking paradigm

The aim of this study was to investigate unconscious priming by the use of a spatial mirror-masking paradigm. Words and nonwords with no under-length letters are mirrored at their horizontal axis. The results are figures of geometric-like forms that contain letters in their upper part. In the three experiments reported in this study, a priming procedure used such mirrored words and nonwords as primes. Participants were ignorant of the nature of the construction of the stimuli. Perceptual reports of the participants revealed that they did not realize that words were hidden in the primes. Nevertheless, they showed priming in all three experiments. Priming effects were replicated with prime–target SOAs of between 1 and 3 s. Functional dissociations were found between ignorant and informed participants. Informed groups showed perceptual and semantic priming, while ignorant groups showed only perceptual priming.

Impacts of a word-picture training on literacy skills in elementary school children and youths with intellectual disabilities

Objective: There is convincing evidence that phonological, orthographic and semantic processes influence children’s ability to learn to read and spell words. So far only a few studies investigated the influence of implicit learning in literacy skills. Children are sensitive to the statistics of their learning environment. By frequent reading they acquire implicit knowledge about the frequency of letter patterns in written words, and they use this knowledge during reading and spelling. Additionally, semantic connections facilitate to storing of words in memory. Thus, the aim of the intervention study was to implement a word-picture training which is based on statistical and semantic learning. Furthermore, we aimed at examining the training effects in reading and spelling in comparison to an auditory-visual matching training and a working memory training program. Participants and Methods: One hundred and thirty-two children aged between 8 and 11 years participated in training in three weekly session of 12 minutes over 8 weeks, and completed other assessments of reading, spelling, working memory and intelligence before and after training. Results: Results revealed in general that the word-picture training and the auditory-visual matching training led to substantial gains in reading and spelling performance in comparison to the working-memory training. Although both children with and without learning difficulties profited in their reading and spelling after the word-picture training, the training program led to differential effects for the two groups. After the word-picture training on the one hand, children with learning difficulties profited more in spelling as children without learning difficulties, on the other hand, children without learning difficulties benefit more in word comprehension. Conclusions: These findings highlight the need for frequent reading trainings with semantic connections in order to support the acquisition of literacy skills.