Significance tests of automatic machine translation evaluation metrics

Automatic evaluation metrics for Machine Translation (MT) systems, such as BLEU, METEOR and the related NIST metric, are becoming increasingly important in MT research and development. This paper presents a significance test-driven comparison of n-gram-based automatic MT evaluation metrics. Statistical significance tests use bootstrapping methods to estimate the reliability of automatic machine translation evaluations. Based on this reliability estimation, we study the characteristics of different MT evaluation metrics and how to construct reliable and efficient evaluation suites.     

From the specification to the implementation of norms: an automatic approach to generate rules from norms to govern the behavior of agents

Open multi-agent systems composed of heterogeneous, autonomous and independently designed agents are usually governed by a set of norms. The established norms regulate the behavior of the agents by pointing out their permissions, prohibitions and obligations. This paper presents a normative language to specify norms and proposes the implementation of such norms by using a rule-based system. The implementation is achieved by automatically transforming the specification of each norm of the system into a set of rules used to govern the behavior of the agents according to the norm. The governance system is able to activate and deactivate norms, to point out the norms violations and fulfillments and to inform about punishments and rewards.     

A relaxation approach to splitting in an automatic theorem prover

The splitting of a problem into subproblems often involves the same variable appearing in more than one of the subproblems. This makes these subproblems dependent upon one another since a solution to one may not qualify as a solution to another. A two stage method of splitting is described which first obtains solutions by relaxing the dependency requirement and then attempts to reconcile solutions to different subproblems. The method has been realized as part of an automatic theorem proper programmed in lisp which takes advantage of the procedural power that lisp provides. The program has had success with cryptarithmetic problems, problems from the blocks world, and has been used as a subroutine in a plane geometry theorem prover.     

An efficient approach to an automatic detection of erythemato-squamous diseases

This paper presents a novel approach to automatic detection of the erythemato-squamous diseases based on fuzzy extreme learning machine (FELM). Enormous computational efforts are required to classify these erythemato-squamous diseases. Some of the approaches performed previously are through fuzzy logic, artificial neural networks and neuro-fuzzy models. FELM-based differential diagnosis of these diseases involves decisions made by fuzzy logic and extreme learning machine (ELM) with greater efficiency in both time and accuracy. In this paper, we develop a user-friendly interface and this tool will be useful for a dermatologist to estimate the six types of erythemato-squamous diseases with the help of patient’s histopathological and clinical data. Then, the developed interface is derived inbuilt using neural networks, adaptive neuro-fuzzy inference system and FELM. A dataset containing records of 366 patients with 34 features that define six disease characteristics was taken, of which 310 records were used as training data and 56 other records used as testing data. The dataset was preprocessed to obtain fuzzy values as input to get more accurate results in FELM. Given a training set of such records, ELM approach is applied. By combining fuzzy logic and ELM, more accurate results with increased performance are obtained with less computational efforts. Finally, the proposed FELM model proves to be a potential solution for the diagnosis of erythemato-squamous diseases with significant improvement in computational time and accuracy compared with other models discussed in the recent literature.     

A computational approach to the variations in Arabic verbal orthography

This paper introduces a finite-state computational approach to four main types of orthographic variations in Arabic which are: the variations in the writing of verb-initial glottal stop, verb-medial glottal stop, verb-final glottal stop, and verb-final weak letters. This approach is based on a linguistically motivated account for 42 types of variations. This account captures generalizations governing such variations, and it is based on syllabification.     

Empirical validation of an automatic usability evaluation method

Today, the success of a software application strongly depends on the usability of its interface, so the evaluation of interfaces has become a crucial aspect of software engineering. It is recognized that automatic tools for graphical user interface evaluation may greatly reduce the costs of traditional activities performed during expert evaluation or user testing in order to estimate the success probability of an application. However, automatic methods need to be empirically validated in order to prove their effectiveness with respect to the attributes they are supposed to evaluate.In this work, we empirically validate a usability evaluation method conceived to assess consistency aspects of a GUI with no need to analyze the back-end. We demonstrate the validity of the approach by means of a comparative experimental study, where four web sites and a stand-alone interactive application are analyzed and the results compared to those of a human-based usability evaluation. The analysis of the results and the statistical correlation between the tool׳s rating and humans׳ average ratings show that the proposed methodology can indeed be a useful complement to standard techniques of usability evaluation.     

On developing an automatic threshold applied to feature selection ensembles

Feature selection ensemble methods are a recent approach aiming at adding diversity in sets of selected features, improving performance and obtaining more robust and stable results. However, using an ensemble introduces the need for an aggregation step to combine all the output methods that confirm the ensemble. Besides, when trying to improve computational efficiency, ranking methods that order all initial features are preferred, and so an additional thresholding step is also mandatory. In this work two different ensemble designs based on ranking methods are described. The main difference between them is the order in which the combination and thresholding steps are performed. In addition, a new automatic threshold based on the combination of three data complexity measures is proposed and compared with traditional thresholding approaches based on retaining a fixed percentage of features. The behavior of these methods was tested, according to the SVM classification accuracy, with satisfactory results, for three different scenarios: synthetic datasets and two types of real datasets (where sample size is much higher than feature size, and where feature size is much higher than sample size).     

Numerical evaluation of the asymptotic energy behavior of intermediate shells with application to two classical benchmark tests

We consider the class of shell problems which are neither purely bending neither purely membrane dominated. In such cases the asymptotic energy norm behavior (which is useful not only because it represents the structure stiffness, but also for numerical comparison purposes) is not a priori known. In this work we apply a numerical procedure in order to estimate the energy behavior of a general shell problem. In order to test its reliability, the method is applied to various problems for which the theoretical energy behavior is known and the results can be compared. Among the problems tested, we have two classical engineering shell benchmarks which are neither bending neither membrane dominated, and for which an analytical evaluation has been obtained in a recent work. All the energy behavior estimates obtained with the numerical method are in perfect agreement with the theoretical values.     

On an approach to designing control of the distributed-parameter processes

Consideration was given to a class of the problems of optimal control of rod (plate) heating by controlling the furnace temperature. Control relies on the process information feedbacked continuously or discretely only from the individual points on the rod where the temperature sensors are installed. For the continuous and discrete observation, the mathematical model of the controlled process was reduced to the pointwise loaded parabolic equation. The paper established formulas for the functional gradient, proposed schemes of their numerical solution, and presented the results of the numerical experiments.     

On an approach to clustering of network traffic

The problem of clustering with a new generalized performance criterion is considered, the concept of the “center of the cluster” is introduced, and it is shown that the definition of the concept is well-defined. Necessary conditions for minimization of the functional are derived in a theorem which encompasses both fuzzy and crisp partitions into clusters. The k-means algorithm, which is based on this necessary condition, finds the optimal cluster iteratively.     

Content annotation for the semantic web: an automatic web-based approach

Semantic Annotation is required to add machine-readable content to natural language text. A global initiative such as the Semantic Web directly depends on the annotation of massive amounts of textual Web resources. However, considering the amount of those resources, a manual semantic annotation of their contents is neither feasible nor scalable. In this paper we introduce a methodology to partially annotate textual content of Web resources in an automatic and unsupervised way. It uses several well-established learning techniques and heuristics to discover relevant entities in text and to associate them to classes of an input ontology by means of linguistic patterns. It also relies on the Web information distribution to assess the degree of semantic co-relation between entities and classes of the input domain ontology. Special efforts have been put in minimizing the amount of Web accesses required to evaluate entities in order to ensure the scalability of the approach. A manual evaluation has been carried out to test the methodology for several domains showing promising results.     

An approach to the automatic design of multiple classifier systems

Multiple classifier systems (MCSs) based on the combination of outputs of a set of different classifiers have been proposed in the field of pattern recognition as a method for the development of high performance classification systems. Previous work clearly showed that multiple classifier systems are effective only if the classifiers forming them are accurate and make different errors. Therefore, the fundamental need for methods aimed to design “accurate and diverse” classifiers is currently acknowledged. In this paper, an approach to the automatic design of multiple classifier systems is proposed. Given an initial large set of classifiers, our approach is aimed at selecting the subset made up of the most accurate and diverse classifiers. A proof of the optimality of the proposed design approach is given. Reported results on the classification of multisensor remote sensing images show that this approach allows the design of effective multiple classifier systems.     

The automatic creation of OLAP cube using an MDA approach

The Model‐Driven Architecture (MDA) is an approach that aligns modeling and automation for software development. By applying such an approach to data warehouse (DW) projects, we can minimize a great deal of time and cost. Furthermore, most of OnLine Analytical Processing (OLAP) platforms seem to be like black boxes that provide wizards only to business intelligence developers to create and manipulate OLAP objects without allowing their sustainability and migration from a platform to another. That is why many works in the literature have proposed using the MDA approach in DW projects. However, most of them have mainly focused on the generation of the DW relational model from the conceptual one, and they overlooked the OLAP model and the cube implementation. To deal with this problem, we propose in this paper an MDA solution to automate the process of getting OLAP cube and its implementation through a set of metamodels and automatic transformations among them. In fact, the proposal generates the OLAP and DW relational models (PSMs) from the conceptual one, using also a PDM model that describes the target business intelligence platform. After that, the source code to create the cube is got from both PSM models. For this aim, we define a set of transformation rules implemented using the Atlas transformation language. Finally, a case study will be provided to validate our approach.     

An approach to automatic adaptation of assembly models

Adaptation plays a fundamental role in case-based design. However, after decades of efforts, automatic adaptation is still an open issue. In works of case-based design, a designer usually chooses a start-up product model (a candidate model) of moderate complexity based on a query model possessing primary new design requirements (kinematic semantics and geometry), then achieves the target design by adapting the candidate model according to the new design requirements and human interventions are often indispensable. To smartly adapt the candidate model to fit the new design requirements, a novel approach to automatic adaptation of assembly models is proposed in this paper. First, in order to effectively identify the corresponding links and interfaces between two non-preregistered assembly models as relevant elements, an attributed kinematic graph is put forward and adopted. Second, based on the attributed kinematic graph, the kinematic semantics of the candidate model is automatically adapted to that of the query model. Third, through performing interface layout transferring, the geometry of the candidate model is automatically adapted to that of the query model based on the corresponding links and interfaces. A prototype system is also implemented to verify the effectiveness of the proposed approach.     

New approach to the automatic segmentation of coronary artery in X-ray angiograms

For the segmentation of X-ray angiograms （XRA）, the essential feature and the prior knowledge of angiographic image were analyzed, and a multi-feature based fuzzy recognition （MFFR） algorithm was proposed to infer the local vessel structure in this paper. Guided by the prior knowledge of artery vessel, a probability tracking operator （PTO） can rapidly track along the artery tree, and walk across the weak region or gaps because of disturbance or preprocessing to angiographic image. Another, the accurate measurement of the vascular axis-lines and diameters can be synchronously implemented in the tracking process. To correctly evaluate the proposed method, a simulated image of CAT and some clinical XRA images were used in the experimentations. The algorithms performed better than the conventional one： given one start-point, on average 92.7% of the visible segments or branches was automatically delineated; the correctness ratio of vessel structure inference reached to 90.0% on the average.     

An approach to the problem of regional accent in automatic speech recognition

An approach to the problem of inter-speaker variability in automatic speech recognition is described which exploits systematic vowel differences in a two-stage process of adaptation to individual speaker characteristics. In stage one, an accent identification procedure selects one of four gross regional English accents on the basis of vowel quality differences within four calibration sentences. In stage two, an adjustment procedure shifts the regional reference vowel space onto the speaker's vowel space as calculated from the accent identification data. Results for 58 speakers from the four regional accent areas are presented.