...Evaluating Commercial Spoken Language Translation Software Harold SOMERS and Yuri SUGITA1 Centre for Computational Linguistics UMIST, PO Box 88 Manchester M60 1QD, England Harold.Somers@umist.ac.uk, sugita_yuri@yahoo.co.jp Abstract1 While spoken language translation remains a research goal, a crude form of it is widely available commercially for Japanese–English as a pipeline concatenation of speech-to-text recognition (SR), text-to-text translation (MT) and text-to-speech synthesis (SS). This paper proposes and illustrates an evaluation methodology for this noisy channel which tries to quantify the relative amount of degradation in translation quality due to each of the contributing modules. A small pilot experiment involving word-accuracy rate for the SR, and a fidelity evaluation for the MT and SS modules is proposed in which subjects are asked to paraphrase translated and/or synthesised sentences from a tourist’s phrasebook. Results show (as expected) that MT is the “noisiest” channel, with SS contributing least noise. The concatenation of the three channels is worse than could be predicted from the performance of each as individual tasks. 1. Introduction Evaluation is without doubt a major aspect of language engineering, including Machine Translation (MT). Although it is still true that no consensus exists regarding the best way to evaluate software, there is general agreement about some of the factors that must be taken into account when deciding what form an evaluation...
Words: 5225 - Pages: 21
...eKNOW 2014 : The Sixth International Conference on Information, Process, and Knowledge Management Malay Semantic Text Processing Engine Benjamin Chu Min Xian MIMOS Berhad Kuala Lumpur, Malaysia mx.chu@mimos.my Liu Qiang MIMOS Berhad Kuala Lumpur, Malaysia qiang.liu@mimos.my Rohana Mahmud University of Malaya Kuala Lumpur, Malaysia rohanamahmud@um.edu.my Arun Anand Sadanandan MIMOS Berhad Kuala Lumpur, Malaysia arun.anand@mimos.my Kow Weng Onn MIMOS Berhad Kuala Lumpur, Malaysia kwonn@mimos.my Dickson Lukose MIMOS Berhad Kuala Lumpur, Malaysia dickson.lukose@mimos.my Abstract—Semantic Text Understanding is a process that transforms text into conceptual representation. In this paper, we propose a Text Understanding System for Malay Language. The system comprises of two components: Morphology Analyzer and Semantic Text Interpreter. Some initial evaluation experiments were conducted on these components to gain explanatory insights into its performance. All the current text processing systems we reviewed are focused on preliminary algorithms and rules associated to lexical, morphological and syntax analysis. In our paper, we developed an integrated approach for a text understanding system that has the ability to represent the semantics of the text. Keywords-Natural Language Understanding; Morphology Interpretation. Processing; Semantic Analysis; Semantic Text Text I. INTRODUCTION The development of fast algorithms to understand and exploit the...
Words: 3718 - Pages: 15
...Dear ***, I'm writing to you in my capacity as Program Committee Co-Chair for the 38th Annual Meeting of the Association for Computational Linguistics, to be held in Hong Kong from October 1--8, 2000. As you may know, the ACL conference is the premier international conference of note in the field of computational linguistics and natural language processing; I am writing to ask whether you would be willing to present a talk at the conference as an invited speaker. Invited talks will be one hour long, including a 10 minute question-answer session. We have not yet established on which day your talk would be scheduled; should you accept this invitation, there is some flexibility we can use to accommodate your own scheduling preferences (although it would be on one of the main conference session days, Tuesday October 3 through Friday October 6). In appreciation of your agreement to provide an invited talk, ACL would provide the cost of an economy class airfare from your home institution to the conference, hotel accommodations during the conference, and free registration to the conference. I will be away for an extended period of time and will not be able to read my email on a regular basis during this time. So please cc Professor Martha Palmer, an area chair and member of the ACL-2000 program committee, in your response. She has kindly agreed to coordinate the invited speaker sessions during my absence. I do very much hope that you will be able to accept this invitation...
Words: 258 - Pages: 2
...What is Forensic Linguistics? According to expert Dr. Carole E. Chaski forensic linguistics applies the theories, constructs and analytical methods of linguistics to questions which arise in civil, criminal and security investigations and adjudication. Dr. Chaski notes that an informed understanding of forensic-linguistics requires familiarity with the broader application of linguistics as a social science. This is because virtually every linguistic nuance has been examined within a legal context. These include: Phonetics & phonology - the system of sounds in human language Morphology - the minimal units of sounds which convey meanings Syntax - the ways in which words can combine into phrases Semantics & Syntax - word, phrasal and sentential meanings Pragmatics - language use in conversation and other forms of discourse Language change - the ways in which language evolves over time Psycholinguistics - how the human cognitive system processes language Neurolinguistics - human neuroanatomy and neuronal functioning Sociolinguistics - identification of linguistic patterns constrained by social, class, racial and gender boundaries Dialectology - identification of dialects related to regional boundaries computational linguistics - tools through which computers can be used to perform linguistic tasks While there are many primary and secondary linguistic subfields, Dr Chaski makes the very important point that one should not be...
Words: 513 - Pages: 3
...linguistics lɪŋgwɪstɪks/ noun 1. 1. the scientific study of language and its structure, including the study of grammar, syntax, and phonetics. Specific branches of linguistics include sociolinguistics, dialectology, psycholinguistics, computational linguistics, comparative linguistics, and structural linguistics. Discourse[edit] A discourse is a way of speaking that emerges within a certain social setting and is based on a certain subject matter. A particular discourse becomes a language variety when it is used in this way for a particular purpose, and is referred to as a register.[25] There may be certain lexical additions (new words) that are brought into play because of the expertise of the community of people within a certain domain of specialisation. Registers and discourses therefore differentiate themselves through the use of vocabulary, and at times through the use of style too. People in the medical fraternity, for example, may use some medical terminology in their communication that is specialised to the field of medicine. This is often referred to as being part of the "medical discourse", and so on. Dialect[edit] A dialect is a variety of language that is characteristic of a particular group among the language speakers.[26] The group of people who are the speakers of a dialect are usually bound to each other by social identity. This is what differentiates a dialect from a register or adiscourse, where in the latter case, cultural identity does not always...
Words: 18394 - Pages: 74
...Introduction This paper will be discussing the information systems and software applications available for use in the accounting, marketing, and human resources departments that work within the constructs of an organization. It will discuss the selected applications, organizational departments, and how they relate to each other. In today’s business environment most if not all businesses rely heavily on computer technology, such as information systems and software applications. Many of these programs have made business faster, more reliable, and much more efficient than in the old days of everything being writen by hand written on paper by a person. Computer based business information systems are not the only form of information system, but keeping with the times computer information systems will be mainly discussed in this essay. A major player in software applications for human resources is PeopleSoft. The PeopleSoft application is designed to help with everyday production, find training tools, and locate documents. It also helps with business performance, and increased productivity due to its many features. The application has a vast number of features and updates including 200 industry specific updates. ("Oracle.com", 2011). Quick books, is a type of business accounting software. Quick books, is mostly used in the small business arena. It has a variety of accounting solutions such as, remote access, remote payroll and outsourcing, online banking and reconciliation...
Words: 291 - Pages: 2
...Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications Computer Applications...
Words: 432 - Pages: 2
...Answer the following questions very briefly (3-4 sentences) IN YOUR OWN WORDS. 1. What is crosstalk? How is it minimized in the case to the twisted pair of wire? My responses to this questions will essentially be generated from my own basic understanding of the reading materials that I found in the unit and this is the concept I had from the textbook that I have seen and researched about this topic that interpreted into my own words. so I have tried this two basic element clarification straight to the point which I understood: 1. Crosstalk refers to the picking up of electromagnetic signals from another adjacent wire by electromagnetic reason. 2. While a pair of wires is twisted together, the electromagnetic signals caused by the two wires eliminate each other as these are of opposing polarity. This serves to decrease the sensitivity of interruption to the nearby wires. 2. Why are two separate frequencies used for uplink and downlink transmission in case of satellite communication? my answer to this question as follows: Two part groups frequencies are used so that one cannot contend with the other and full duplex communication is possible. Also, the different idea is that the Power needed to transfer a signal is equal to the frequency of the signal. Plus more power demand more would be the power of the system. As there are constraints on the load that can be carried with the satellite, mainly downlinking frequency is lower than the up linking one. 3. What is the...
Words: 1040 - Pages: 5
...TACIT VERSUS CONSCIOUS KNOWLEDGE * Tacit vs. conscious knowledge * Descriptive vs. prescriptive rules * Each human language corresponds to a highly complex system of rules * Linguistics involves studying these rules and their implications SPEECH AND PHONEMES * The Language Faculty is a mental system that converts a physical signal (sound or sign) into meaning, and vice versa. * The Language Faculty makes several intermediate step between sound and meaning. * Phonemes are the smallest segment of speech that leads to meaningful contrasts between words is. * The properties of the vocal folds (the source), such as frequency and amplitude, are not critical to speech sounds. * It is the properties of the vocal tract (the filter) that are critical. * Formants are the highest amplitude peaks in the frequency spectrum created by the human vocal tract. Different formants lead to different speech sounds. ARTICULORY FEATURES AND CATEGORICAL PERCEPTION * Formants are the physical properties of phonemes that appear to be most critical to speech perception * However, we don’t use formants directly (we know this because of the three problems with formants). * Instead, during comprehension we convert formants into the articulatory features that the speaker used to create the formants. * Consonants are each made up of three features: place of articulation, manner of articulation, and voicing. * Vowels are each made up of two features:...
Words: 411 - Pages: 2
...The theoretical framework for teachers’ beliefs in this study is based on a computational model of second language acquisition proposed by Ellis (1998) that highlights four major stages that relates roughly to common beliefs on the process of how grammar should be taught. The first stage, input, stresses on the materials and the modification of input used in teaching grammar in the classroom. The second, explicit knowledge, provides explicit instruction; the third, output, emphasizes production practice; and the final stage, negative feedback, highlights the role of error correction. The model can be seen in Figure 1 with the four stages highlighted by the letters A, B, C, and D respectively. Figure _: A Computational Model of L2 Acquisition...
Words: 776 - Pages: 4
...Why Ambiguity? Ariel Cohen Ben-Gurion University arikc@bgu.ac.il All human languages are ambiguous. Ambiguity is not restricted to some special constructions that linguists are fond of discussing, but is quite ubiquitous. It is hard, in fact, to find a sentence that is not ambiguous. This fact is all too familiar to computational linguists: “One often hears in computational linguistics about completely unremarkable sentences with hundreds of parses, and that is in fact no exaggeration.” (Abney, 1996). The question is, simply, why? Why is language ambiguous? Krifka (2002) raises a similar question concerning vagueness. He shows that there are good reasons for language to allow, and even encourage vagueness. Vagueness, however, is not ambiguity. When we are told that the theater is far from here, we may be unsure as to the precise distance; but we know what the speaker intends, and we can draw inferences based on this, e.g., that we should take a cab instead of walking to the theater. But with an ambiguous term, the intended meaning is not merely insufficiently specified; it is not known, until the term is disambiguated. When we hear that John has a kid, we draw very different inferences if John is a father or if he owns a young goat. Thus, the advantages of vagueness do not seem to apply to ambiguity, and we are back to the question: why ambiguity? Some may say that it doesn’t matter. Humans possess very powerful mechanisms for disambiguation; these mechanisms resolve ambiguities,...
Words: 1571 - Pages: 7
...Simply defined, linguistics is the scientific study of language. Though various types of language studies (including grammar and rhetoric) can be traced back over 2,500 years, the era of modern linguistics is barely two centuries old. Kicked off by the late-18th-century discovery that many European and Asian languages descended from a common tongue (Proto-Indo-European), modern linguistics was reshaped, first, by Ferdinand de Saussure (1857-1913) and more recently by Noam Chomsky (born 1928). The systematic study of the nature, structure, and variation of language. Major subfields of linguistics include phonetics, phonology, morphology, syntax, semantics, pragmatics, and discourse analysis. The founder of modern structural linguistics was Ferdinand de Saussure (1857-1913), whose most influential work, Course in General Linguistics, was edited by his students and published in 1916. Source: An Introduction to Language by Victoria Fromkin and Robert Rodman, 6th Ed.) Part One: Introduction to Linguistics Every human knows at least one language, spoken or signed. Linguistics is the science of language, including the sounds, words, and grammar rules. Words in languages are finite, but sentences are not. It is this creative aspect of human language that sets it apart from animal languages, which are essentially responses to stimuli. The rules of a language, also called grammar, are learned as one acquires a language. These rules include phonology, the sound system, morphology, the...
Words: 10632 - Pages: 43
...CHAPTER 1 INTRODUCTION 1.1 Background of the Study "Language and Media" is a comprehensive introduction to how language interacts with media. It investigates the forms of language found in media discourse; how patterns in such language use contribute to recognizable media genres and styles; and, broader social themes and consequences that arise from media language. It uses a wide variety of real texts from the media specifically from Magazines. Boles (2012) believes that each magazine uses linguistic techniques these language features are creatively used in magazines by the writers to attract and entertain the target audience as they only have a small space and time to get the attention of the potential buyer of the product. Bell (2010) emphasizes the importance of the processes which produce media language, as stories are molded and modified by various hands. He stresses it is indeed stories that journalists and editors produce, not articles. These stories have viewpoint, values and structure that can be analyzed. He is concerned, too, with the role of the audience in influencing media language styles, and in understanding, forgetting or misconceiving the news presented to it. A feature story is a special human interest story article that is not closely tied to a recent news event. It focuses on particular people, places, and events, and it goes into great detail regarding concepts and ideas of specific market interest. The study primarily aims to provide...
Words: 1649 - Pages: 7
...Current Events in Business Research Proper research in business is vital to its success. This paper will discuss how well the article Computer Modelling and Simulation of Bakeries' Production Planning written and researched by Walid Barakat Hussein, Florian Mitzscherling Hecker, Thomas Martin Becker published in 2009. The article summarizes the efforts of bakeries selling ten different products using the ARENA computer simulation system to streamline production and prevent “bottle-necking”. The first step of effective business research is stating a clear purpose. The article starts out with an introduction that states businesses past troubles with understanding the “sequence of activities required to complete a process-whether fulfilling an order, producing a part, or servicing a customer,” (par 1) which is what the ARENA computer system is supposed to help with. Later in the introduction: “This research work presents a methodology of modeling and simulation of production system using Arena software simulation tool, in order to deal efficiently with the production procedures and processes, and propose an appropriate system structure that will operate according to given specifications,” (par 3). The introduction stated a very clear purpose, which helps those reading the article as well as following the first step of proper business research. The research process was thoroughly detailed and the design was thoroughly planned. In the method and case study section on the article...
Words: 496 - Pages: 2
...44325-3909, USA ABSTRACT Several classes of computational methods are available for computer simulation of electromagnetic wave propagation and scattering at optical frequencies: Discrete Dipole Approximation, the T-matrix − Extended Boundary Condition methods, the Multiple Multipole Method, Finite Difference (FD) and Finite Element (FE) methods in the time and frequency domain, and others. The paper briefly reviews the relative advantages and disadvantages of these simulation tools and contributes to the development of FD methods. One powerful tool – FE analysis − is applied to optimization of plasmon-enhanced AFM tips in apertureless near-field optical microscopy. Another tool is a new FD calculus of “Flexible Local Approximation MEthods” (FLAME). In this calculus, any desirable local approximations (e.g. scalar and vector spherical harmonics, Bessel functions, plane waves, etc.) are seamlessly incorporated into FD schemes. The notorious ‘staircase’ effect for slanted and curved boundaries on a Cartesian grid is in many cases eliminated – not because the boundary is approximated geometrically on a fine grid but because the solution is approximated algebraically by suitable basis functions. Illustrative examples include problems with plasmon nanoparticles and a photonic crystal with a waveguide bend; FLAME achieves orders of magnitude higher accuracy than the standard FD methods, and even than FEM. Keywords: wave propagation, computational methods, flexible approximation, photonic...
Words: 5791 - Pages: 24
