拼写纠正在拼音输入法中的应用

中文输入法一直是中文语言研究的一个难题,文中以拼音整句输入法为基础,提出了在中文输入过程中的拼写自动修改,通过对用户输入过程中所犯各种错误的分析,建立了一种有效可行的打字模型,通过收集用户真实输入的数据,统计得到用户的打字模型的参数;同时基于大量的中文文本,训练得到一个强大的中文语言模型,并与中文的打字模型相结合,采用类似语音识别的技术,修改用户输入中的各种错误,并得到最适合的汉字。同时,拼写纠正不仅可以进行用户自适应,而且还适用于各种语言。     

The effects of work pace on within-participant and between-participant keying force,electromyography, and fatigue

A laboratory study was conducted to determine the effects of work pace on typing force, electromyographic (EMG) activity, and subjective discomfort. We found that as participants typed faster, their typing force and finger flexor and extensor EMG activity increased linearly. There was also an increase in subjective discomfort, with a sharp threshold between participants' self-selected pace and their maximum typing speed. The results suggest that participants self-select a typing pace that maximizes typing speed and minimizes discomfort. The fastest typists did not produce significantly more finger flexor EMG activity but did produce proportionately less finger extensor EMG activity compared with the slower typists. We hypothesize that fast typists may use different muscle recruitment patterns that allow them to be more efficient than slower typists at striking the keys. In addition, faster typists do not experience more discomfort than slow typists. These findings show that the relative pace of typing is more important than actual typing speed with regard to discomfort and muscle activity. These results suggest that typists may benefit from skill training to increase maximum typing speed. Potential applications of this research includes skill training for typist.     

Effects of computer monitor viewing angle and related factors on strain, performance, and preference outcomes

A model of visual and musculoskeletal strain associated with computer monitor placement was developed. The main premise of which is that monitor placement decisions must take into consideration development of both visual and musculoskeletal strains. Certain factors in the model that were thought to affect one or both types of strain. or that were considered important to rule out for effect, were tested in a lab setting. These factors were viewing angle (eye level, midlevel, low level), monitor size (14 in., 19 in.), keyboard familiarity (touch typist, nontouch typist), and task (reading, mousing, typing). Outcomes included indicators of visual and musculoskeletal strain, preference, and performance. Muscle activity was generally greater for the low viewing angle, for the standard monitor (14 in.), and for non-touch typists. Participants preferred the midlevel placement. Task performance was slightly diminished with eye-level placement. Results are interpreted in relation to the model and to several hypotheses that were formed to focus the inquiry. Actual or potential applications of this research include monitor placement decisions in the design or modification of computer workstations.     

Mobile text-entry and visual demands: reusing and optimizing current solutions

Mobile devices are increasingly used for text-entry in contexts where visual attention is fragmented and graphical information is inadequate, yet the current solutions to typing on virtual keyboards make it a visually demanding task. This work looks at assistive technologies and interface attributes as tools to ease the task. Two within-subject experiments were performed with 23 and 17 participants, respectively. The first experiment aimed to understand how walking affected text-entry performance and additionally to assess how effective assistive technologies can be in mobile contexts. In the second experiment, adaptive keyboards featuring character prediction and pre-attentive attributes to ease visual demands of text-entry interfaces were developed and evaluated. It has been found that both text-input speed and overall quality are affected in mobile situations. Contrary to the expectations, assistive technologies proved ineffective with visual feedback. The second experiment showed that pre-attentive attributes do not affect users’ performance in task-entry tasks, even though a 3.3–4.3 % decrease in error rates was measured. It was found that users reduce walking speed to compensate for challenges placed by mobile text-entry. Caution should be exercised when transferring assistive technologies to mobile contexts, since they need adaptations to address mobile users’ needs. Also, while pre-attentive attributes seemingly have no effect on experienced QWERTY typists’ performance, they showed promise for both novice users and typists in attention-demanding contexts.     

Improving static audio keystroke analysis by score fusion of acoustic and timing data

In this paper we investigate the capacity of sound & timing information during typing of a password for the user identification and authentication task. The novelty of this paper lies in the comparison of performance between improved timing-based and audio-based keystroke dynamics analysis and the fusion for the keystroke authentication. We collected data of 50 people typing the same given password 100 times, divided into 4 sessions of 25 typings and tested how well the system could recognize the correct typist. Using fusion of timing (9.73%) and audio calibration scores (8.99%) described in the paper we achieved 4.65% EER (Equal Error Rate) for the authentication task. The results show the potential of using Audio Keystroke Dynamics information as a way to authenticate or identify users during log-on.     

A decomposition theorem for hyper-algebraic extensions of language families

In modern theories of rewriting structures, hyper-sentential and hyper-algebraic extensions of languages-families have abstracted the imminent features of iterated parallel substitution. After introducing the concept of a (depth-bounded) translation, we show that for each language L hyper-algebraic over a natural family $\text{F}$ there are $\text{F}$-translations Δ, ?D and languages L₁, …, L_m hyper-sentential over $\text{F}$ such that

, for some p, q ? 0.Two specializations of this result are given, when more assumptions are made about $\text{F}$. These are, firstly, a translation theorem and, secondly, an alphabetic homomorphism theorem for hyper-algebraic extensions (an alphabetic homomorphism is a letter-to-letter or letter-to-ε homomorphism).     

Heap-Bounded Assembly Language

We present a first-order linearly typed assembly language, HBAL, that allows the safe reuse of heap space for elements of different types. Linear typing ensures the single pointer property, disallowing aliasing but allowing safe, in-place-update compilation of programming languages. We prove that HBAL is sound for a low-level untyped model of the machine, using a satisfiability relation that captures when a location correctly models a value of some type. This interpretation is closer to the machine than previous abstract machines used for typed assembly language models, and we separate typing of the store from an untyped operational semantics of programs, as would be required for proof-carrying code. Our ultimate aim is to design a family of assembly languages that have high-level typing features for expressing resource-bound constraints. We want to link the assembly-level with high-level languages expressing similar constraints, to provide end-to-end guarantees and a viable framework for proof-carrying code. HBAL is a first exemplifying step in this direction. It is designed as a target low-level language for Hofmann's LFPL language. Programs written in LFPL run in a bounded amount of heap space, and this property carries over when they are compiled to HBAL: the resulting program does not allocate store or assume an external garbage collector. Following LFPL, we include a special diamond resource type that stands for a unit of heap space of uncommitted type.     

Cervicobrachial muscle response to cognitive load in a dual-task scenario

People working in an office environment often have to deal with significant cognitive workload due to the coordination of multiple, simultaneous tasks. The objective of this research was to examine the impact of cognitive load in office-type tasks on physical-stress response, using a dual-task paradigm involving a primary cognitive task and secondary typing task. The central hypothesis of this research was that altering the demands of the cognitive task would lead to a difference in physical stress-level and performance. Cognitive load was manipulated by presenting participants with three different types of cognitive tasks described in Rasmussen's () taxonomy, including skill-, rule-, and knowledge-based tasks. Dependent variables examined in the study included: (1) electromyographic activity of the upper trapezius (pars descendens) and cervical erector spinae muscles, (2) performance in a secondary typing task, and (3) subjective measures of stress and cognitive workload. The results of this study revealed that the primary task causing the highest level of perceived workload also produced 61% higher muscle activity in the right trapezius, and 6 and 11% higher activity in the left and right cervical erector spinae, respectively, in comparison to muscle activity associated with the cognitive task causing the lowest perceived workload. With respect to performance, a 23% decrease was observed in typing productivity when the rule-based task was completed simultaneously vs. typing in the absence of any additional cognitive task (the baseline condition). This information may be used to better organize work activities in office environments to increase performance and reduce stress.     

Type Destructors

We study a variant of System F_≤ that integrates and generalizes several existing proposals for calculi with “structural typing rules.” To the usual type constructors (→, ×, All, Some, Rec) we add a number of type destructors, each internalizing a useful fact about the subtyping relation. For example, in F_≤ with products every closed subtype of a product S×T must itself be a product S′×T′ with S′<:S and T′<:T. We internalise this observation by introducing type destructors .1 and .2 and postulating an equivalence T=_ηT.1×T.2 whenever T<:U×V (including, for example, when T is a variable). In other words, every subtype of a product type literally is a product type, modulo η-conversion. Adding type destructors provides a clean solution to the problem of polymorphic update without introducing new term formers, new forms of polymorphism, or quantification over type operators. We illustrate this by giving elementary presentations of two well-known encodings of objects, one based on recursive record types and the other based on existential packages. The formulation of type destructors poses some tricky meta-theoretic problems. We discuss two different variants: an “ideal” system where both constructors and destructors appear in general forms, and a more modest system, F_≤^TD, which imposes some restrictions in order to achieve a tractable metatheory. The properties of the latter system are developed in detail.     

Skill-level based examination of forearm muscle activation associated with efficient wrist and finger movements during typing

This study aims to elucidate the relationship between the wrist and finger movements and forearm muscle activation of twelve young people (age: 21.1 ± 0.76 years, nine men and three women) with different typing skills. We hypothesize that skilled typists (STs) could move their wrist and finger joints faster than unskilled typists (UTs) because they could efficiently use their muscles according to the activity characteristics of the flexors and extensors of the wrist joint. We measured wrist and finger movements using a 3D motion capture system and forearm muscle activation using surface electromyography during the typing task. We analyzed the entire task and the time when the U key was entered during the same task. The angular velocity of the wrist and finger flexion/extension and the muscle activation of the wrist flexors was higher in the STs than in the UTs, while the muscle activation of the wrist extensors was higher in the latter than that in the former. Our results showed that STs may have used their forearm muscles to take advantage of the physical characteristics of the keys and the spring characteristics of their muscles and tendons. It was suggested that they placed less mechanical stress on their finger muscles and tendons when pressing and releasing the keys.     

An Extension of System F with Subtyping

System F is a well-known typed λ-calculus with polymorphic types, which provides a basis for polymorphic programming languages. We study an extension of F, called F_<: (pronounced ef-sub), that combines parametric polymorphism with subtyping. The main focus of the paper is the equational theory of F_<:, which is related to PER models and the notion of parametricity. We study some categorical properties of the theory when restricted to closed terms, including interesting categorical isomorphisms. We also investigate proof-theoretical properties, such as the conservativity of typing judgments with respect to F. We demonstrate by a set of examples how a range of constructs may be encoded in F_<:. These include record operations and subtyping hierarchies that are related to features of object-oriented languages.     

Cervicobrachial muscle response to cognitive load in a dual-task scenario

People working in an office environment often have to deal with significant cognitive workload due to the coordination of multiple, simultaneous tasks. The objective of this research was to examine the impact of cognitive load in office-type tasks on physical-stress response, using a dual-task paradigm involving a primary cognitive task and secondary typing task. The central hypothesis of this research was that altering the demands of the cognitive task would lead to a difference in physical stress-level and performance. Cognitive load was manipulated by presenting participants with three different types of cognitive tasks described in taxonomy, including skill-, rule-, and knowledge-based tasks. Dependent variables examined in the study included: (1) electromyographic activity of the upper trapezius (pars descendens) and cervical erector spinae muscles, (2) performance in a secondary typing task, and (3) subjective measures of stress and cognitive workload. The results of this study revealed that the primary task causing the highest level of perceived workload also produced 61% higher muscle activity in the right trapezius, and 6 and 11% higher activity in the left and right cervical erector spinae, respectively, in comparison to muscle activity associated with the cognitive task causing the lowest perceived workload. With respect to performance, a 23% decrease was observed in typing productivity when the rule-based task was completed simultaneously vs. typing in the absence of any additional cognitive task (the baseline condition). This information may be used to better organize work activities in office environments to increase performance and reduce stress.     

L0-The first five years of an automated language acquisition project

The L₀ project at ICSI and UC Berkeley attempts to combine not only vision and natural language modelling, but also learning. The original task was put forward in (Feldman et al. 1990a) as a touchstone task for AI and cognitive science. The task is to build a system that can learn the appropriate fragment of any natural language from sentence-picture pairs. We have not succeeded in building such a system, but we have made considerable progress on component subtasks and this has led in a number of productive and surprising directions.     

Paraphrastic language models

Natural languages are known for their expressive richness. Many sentences can be used to represent the same underlying meaning. Only modelling the observed surface word sequence can result in poor context coverage and generalization, for example, when using n-gram language models (LMs). This paper proposes a novel form of language model, the paraphrastic LM, that addresses these issues. A phrase level paraphrase model statistically learned from standard text data with no semantic annotation is used to generate multiple paraphrase variants. LM probabilities are then estimated by maximizing their marginal probability. Multi-level language models estimated at both the word level and the phrase level are combined. An efficient weighted finite state transducer (WFST) based paraphrase generation approach is also presented. Significant error rate reductions of 0.5–0.6% absolute were obtained over the baseline n-gram LMs on two state-of-the-art recognition tasks for English conversational telephone speech and Mandarin Chinese broadcast speech using a paraphrastic multi-level LM modelling both word and phrase sequences. When it is further combined with word and phrase level feed-forward neural network LMs, a significant error rate reduction of 0.9% absolute (9% relative) and 0.5% absolute (5% relative) were obtained over the baseline n-gram and neural network LMs respectively.     

Effect of physical workload and modality of information presentation on pattern recognition and navigation task performance by high-fit young males

Abstract

Many occupations require both physical exertion and cognitive task performance. Knowledge of any interaction between physical demands and modalities of cognitive task information presentation can provide a basis for optimising performance. This study examined the effect of physical exertion and modality of information presentation on pattern recognition and navigation-related information processing. Results indicated males of equivalent high fitness, between the ages of 18 and 34, rely more on visual cues vs auditory or haptic for pattern recognition when exertion level is high. We found that navigation response time was shorter under low and medium exertion levels as compared to high intensity. Navigation accuracy was lower under high level exertion compared to medium and low levels. In general, findings indicated that use of the haptic modality for cognitive task cueing decreased accuracy in pattern recognition responses.

Practitioner Summary: An examination was conducted on the effect of physical exertion and information presentation modality in pattern recognition and navigation. In occupations requiring information presentation to workers, who are simultaneously performing a physical task, the visual modality appears most effective under high level exertion while haptic cueing degrades performance.     

On purely morphic characterizations of context-free languages

In this paper we show the following: For any λ-free context-free language L there effectively exist a weak coding g, a homomorphism h such that L=gh⁻¹ (∣cD₂), where D₂ is the Dyck set over a two-letter alphabet. As an immediate corollary it follows that for any λ-free context-free language L there exist a weak coding g and a mapping F such that L=gF⁻¹(∣c).     

Mechanizing Metatheory Without Typing Contexts

When mechanizing the metatheory of a programming language, one usually needs many lemmas proving structural properties of typing judgments, such as permutation and weakening. Such structural lemmas are sometimes unnecessary if we eliminate typing contexts by expanding typing judgments into their original hypothetical proofs. This technique of eliminating typing contexts, which has been around since Church (J Symb Log 5(2):56–68, 1940), is based on the view that entailment relations, such as typing judgments, are just syntactic tools for displaying only the hypotheses and conclusion of a hypothetical proof while hiding its internal structure. In this paper, we apply this technique to parts 1A/2A of the textscPoplMark challenge (Aydemir et al. 2005) and experimentally evaluate its efficiency by formalizing System F _?<?:? in the Coq proof assistant in a number of different ways. An analysis of our Coq developments shows that eliminating typing contexts produces a more significant reduction in both the number of lemmas and the count of tactics than the cofinite quantification, one of the most effective ways of simplifying the mechanization involving binders. Our experiment with System F _?<?:? suggests three guidelines to follow when applying the technique of eliminating typing contexts.     

Word co-occurrence features for text classification

In this article we propose a data treatment strategy to generate new discriminative features, called compound-features (or c-features), for the sake of text classification. These c-features are composed by terms that co-occur in documents without any restrictions on order or distance between terms within a document. This strategy precedes the classification task, in order to enhance documents with discriminative c-features. The idea is that, when c-features are used in conjunction with single-features, the ambiguity and noise inherent to their bag-of-words representation are reduced. We use c-features composed of two terms in order to make their usage computationally feasible while improving the classifier effectiveness. We test this approach with several classification algorithms and single-label multi-class text collections. Experimental results demonstrated gains in almost all evaluated scenarios, from the simplest algorithms such as kNN (13% gain in micro-average F₁ in the 20 Newsgroups collection) to the most complex one, the state-of-the-art SVM (10% gain in macro-average F₁ in the collection OHSUMED).     

Efficient evaluation of specific queries in constraint databases

Let F₁,…,Fs∈R[X₁,…,Xn] be polynomials of degree at most d, and suppose that F₁,…,Fs are represented by a division free arithmetic circuit of non-scalar complexity size L. Let A be the arrangement of Rn defined by F₁,…,Fs.For any point x∈Rn, we consider the task of determining the signs of the values F₁(x),…,Fs(x) (sign condition query) and the task of determining the connected component of A to which x belongs (point location query). By an extremely simple reduction to the well-known case where the polynomials F₁,…,Fs are affine linear (i.e., polynomials of degree one), we show first that there exists a database of (possibly enormous) size sO(L+n) which allows the evaluation of the sign condition query using only (Ln)^O(1)log(s) arithmetic operations. The key point of this paper is the proof that this upper bound is almost optimal.By the way, we show that the point location query can be evaluated using dO(n)log(s) arithmetic operations. Based on a different argument, analogous complexity upper-bounds are exhibited with respect to the bit-model in case that F₁,…,Fs belong to Z[X₁,…,Xn] and satisfy a certain natural genericity condition. Mutatis mutandis our upper-bound results may be applied to the sparse and dense representations of F₁,…,Fs.     

Performance effects of reduced proprioceptive feedback on touch typists and casual users in a typing task

This study examined performance and acceptance effects of lack of kinesthetic and tactile feedback from the keyboard in a typing task with two subject groups of differing skill level: touch typists and casual users. Subjects' objective performance (e.g., speed, accuracy, throughput) and subjective acceptance (questionnaire) was evaluated for both a conventional full travel keyboard and a prototype piezoelectric flat keyboard which lacked familiar kinesthetic and tactile feedback. Any performance decrement present with the flat keyboard was expected to diminish with practice for the touch typists due to transfer and adaptation of typing skills. Performance for both subject groups was significantly higher with the conventional keyboard and touch typists' performance was more adversely affected by the flat keyboard than casual users'. No performance improvement with practice was found for one subject group relative to the other or for one keyboard relative to the other. It was concluded the touch typists were unable to adapt to the unusual feedback conditions present with the piezo-electric fiat keyboard.