LINGUIST List 24.399
Tue Jan 22 2013
Review: Applied Linguistics; Computational Linguistics; Semantics: Plastina (2012)
Editor for this issue: Monica Macaulay
Sandra Petroni <sandra.petroni
Applied Linguistics and Semantic Web Apps: Cases of Mediated Discursive Practices
E-mail this message to a friend
Discuss this message
Book announced at http://linguistlist.org/issues/23/23-3114.html
AUTHOR: Anna Franca PlastinaTITLE: Applied Linguistics and Semantic Web Apps: Cases of Mediated Discursive PracticesSERIES TITLE: LINCOM Studies in SemanticsPUBLISHER: Lincom GmbHYEAR: 2012
REVIEWER: Sandra Petroni, University of Roma - Tor Vergata
“Applied Linguistics and Semantic Web Apps” provides a general analysis ofSemantic Web Apps through which new discursive practices are mediated indigital environments. The three case studies presented show how AppliedLinguistics research can be integrated into Semantic Web enquiry. In fact, themain research question posited by the author is: “How would an AppliedLinguist respond to discursive practices mediated through Semantic Web Apps?”(p. 18).
The book includes eight chapters, preceded by a list of figures, tables, and alist of Semantic Web Apps. These chapters are organized in two main parts,preceded in turn by an introduction.
The first part, ‘Sharpening Research in Applied Linguistics: the Role ofSemantic Web Apps’ (pp. 21-73), is composed of three chapters and each chapterincludes an exhaustive and well supported presentation of the topics. Thesecond part, ‘Mediated Discursive Practices on the Semantic Web’ (pp. 75-173),is composed of five chapters. With the exception of Chapters 4 and 8, thechapters provide both the theoretical background and the research issue interms of research methodology, tools, method and procedure; show findings oneach single simulation; offer comments on data triangulation; and conclude bydescribing the implications for mediated practices. The book ends withReferences.
Chapter 1, ‘Applying Linguistics’ (pp. 23-36), traces the evolution of AppliedLinguistics over the past few decades. The author makes reference toWeidemann’s (2003) classification of the six generations of this discipline toillustrate its diverse models and traditions - such as the behaviourist model,linguistic extended paradigm model, multi-disciplinary model, second languageacquisition research, constructivism, and post-modernism - with their specificapproaches and consequential changes. As the author states, thanks to thesixth generation, “the work of applied linguistics [is conceived of] asdirected both towards investigation on language learning/teaching in all itsdifferent facets, and as committed to research in a multi-disciplinaryperspective” (p. 28). Thus, the pivotal aspects of Applied Linguistics are“openness” deriving from “the permeable boundaries of the discipline”;“interdisciplinarity” in terms of “interdisciplinary collaboration andintegration with other different disciplines”; and “real-world enquiry” thatis “its engagement in solving real-world language-based problems” (p. 28). Allthese three facets have made the osmosis between applied linguistics andcomputational linguistics not only possible but also necessary. Apart from thefully-integrated use of computers in language learning/teaching and the riseof new branches, such as Internet Linguistics or Netlinguistics, the authorunderlines that the attention of applied linguists should be directed to thenew emergent semantic web technologies that have been decidedly affectingcomputational linguistics.
Chapter 2, ‘The Semantic Web’ (pp. 37-51), posits two main issues: to motivatethe need to design an ‘intelligent’ Semantic Web and to identify those maincomponents that can have a significant impact on Applied Linguistics. Thefirst issue delineates a topic that has been analysed from different points ofview in the last decades: information overload, or rather, “the overwhelmingamount of unstructured information [that] hinders effective informationretrieval and management” due to the fact that “on the one hand, thisexponential quantity of information is currently intelligible only to humans,while on the other, it is only processable by machines” (p. 51). Afterdepicting the evolution and the mission of the Semantic Web, the second issueidentifies its components, i.e. Natural Language Processing (NLP),machine-based learning and reasoning, and intelligent applications. The authoroffers an exhaustive explanation of these three areas of investigation mostlyrelated to Artificial Intelligence (AI) and Computer Science.
Chapter 3, ‘Semantic Web Apps: Text Mining’ (pp. 53-73), focuses on the roleplayed by Semantic Web Apps, their main purposes of use and operationaleffectiveness. As the author states, “[…], a key feature of all theseapplications is that they attempt to determine the meaning of texts and/or ofother data in order to create connections which are meaningful to users” (p.54). Their functions then are to process and manage information and tofacilitate and optimize information retrieval in terms of time andeffectiveness. In order to meet the last condition, two main factors arenecessary: the availability of ontologies, “frameworks of knowledgerepresentation, describing a specific area of knowledge through the conceptsof the domain, and their properties and relationships” (p. 58); and thepotentialities of semantic search engines that, differently from the commonsearch engines, exploit semantics to generate more appropriate results bydisambiguating polysemy and synonymy. Text mining permits these Semantic WebApps to “semantize” (p. 72) information and to examine meaning at differentlanguage levels (e.g., a Web annotation app for information extraction impliesoperating at the morphological level; a text clouding app for textsummarization works on the semantic level; a sentiment classification app forsentiment analysis looks at the pragmatic level).
Chapter 4, ‘Setting up Research Enquiry’ (pp. 78-95), provides “a detaileddescription of how research enquiry was set up to conduct mediated practiceswith Semantic Web Apps for language-related enquiry” (p. 78). Based on asemantic-web app approach - rather than the NLP approach, the corpus-basedapproach, the Web-as-Corpus approach, and the search engine approach -research enquiry can move beyond the classical fields of computationallinguistics, corpus linguistics and lexicography, can access resourcesdirectly online and make use of open-source tools. The research frameworkutilized here starts from the following research question: “To which extentdoes the use of Semantic Web Apps add new theoretical insight intolanguage-based enquiry in different sub-fields of applied linguistics?” (p.82). Computer simulation is used as research method and qualitative empiricalinvestigation is undertaken as research methodology. The research procedurefollows three phases: Orientation (selection of strands of AppliedLinguistics, refinement of research questions, selection of Semantic WebApps); Focused Exploration (collection of data); Verification (datatriangulation). The rigorous respect that the author has for this researchframework aims at legitimating research processes.
Chapter 5, ‘Enhancing Consumer Health Vocabulary: The Case of Ontology-basedSearches’ (pp. 97-122), focuses on mediated discursive practices grounded inprofessional and specialised domains (health communication in this case). Theauthor hypothesises that, first, an ontology-based approach to health consumerqueries helps users improve their knowledge of specialised terminology, and,second, it offers more effective semantic results than those offered by commonsearch engines (e.g. Google).The theoretical background provided then by theauthor is based on Frame Semantics (Fillmore 1982, 1985) and Frame-BasedTerminology (Faber et al. 2005). Research tools are respectively a samplemedical transcription of a discharge summary, including clinical information,and NLMplus (http://nlmplus.com
), a semantic search engine which make use of60 medical ontologies related to verbal and visual resources. The author showsthat the Focused Exploration phase (selecting ‘obscure’ medical terms andsimulating semantic searches in NLMplus) confirms the first hypothesis, andthe Verification phase (simulating the same queries in Google and comparingdata with results of the previous phase) confirms the second hypothesis.Findings demonstrate that “NLMplus is a powerful Semantic Web App as it iscapable of extracting biomedical concepts across different specializedontologies, ensuring trust and reliability” (p. 113), more than onlinedictionaries or Wikipedia. The author calls attention to some limitations tothis approach, i.e. the user’s digital and medical literacy and his/herEnglish language competence.
Chapter 6, ‘Developing L2 Readers’ Comprehension Strategies: The Case of theText Clouding Approach’ (pp. 123-144), “seeks to understand how learners canmanipulate English L2 texts via Semantic Web Apps in order to develop theircognitive strategies for better reading comprehension” (p. 125). The authorposits two hypotheses: first, Semantic Web Apps based on the text-cloudingtechnique can be used as “effective scaffolds for the development of EFLreaders’ higher-level reading comprehension strategies” (p. 131); second, thistechnique is more effectual than linear post-reading activities. Thetheoretical background here is based, on the one hand, on Dual Coding Theory(Paivio 1986, 2006) that claims that human cognition works simultaneously ontwo subsystems (verbal for language processing and visual for imageryprocessing); on the other hand, on Text Visualization, the transformation ofwritten texts into graphical representations. Research tools are respectivelya sample reading text (Common European Framework B1 level) with fivetraditional post-reading activities and VocabGrabber(www.visualthesaurus.com/vocabgrabber
), an online word cloud generator. TheFocused Exploration phase (learners interact and familiarise themselves withthe learning context; they transform the text into a word cloud; they areengaged in speculating on the visual representation) aims at confirming thefirst hypothesis while the Verification phase (learners analyse the new outputfor comprehension and then they are requested to report on the essentialmeaning of the text) confirms the second one. Findings reveal thattext-clouding supports L2 students in reading the original text by visualizingthe cloud of keywords, in identifying main ideas through the options ofrelevance and occurrence provided by the app, and in expanding schemata. Lastbut not least, as stated by the author, text-clouding develops students’digital literacy.
Chapter 7, ‘Attitude Detection in Hotel Reviews: The Case of SentimentAnalyzers’ (pp. 145-168), deals with mediated discursive practices whereemotional experience plays a strategic role. For this reason, the authorfocuses on the language of emotion and appraisal used in textual practice,such as hotel reviews available online, and on the opportunity to detect itcomputationally. The hypotheses of this enquiry then are: first, sentimentdetection can be effective if the Semantic Web Apps involved are based onAppraisal Theory and in particular on the appraisal Framework; second, thisapproach allows users to discriminate biased or unbiased stances presentrespectively in hotel homepages or reviews. As theoretical background, theauthor refers to Aronson et al. (2005) for Appraisal Theory and to Martin(2000) and Martin & White (2005) for the Appraisal Framework. AffectiveComputing (Picard 1997) is briefly mentioned. Research tools are,respectively, a corpus of customer reviews taken from Booking.com and anonline sentiment analyzer, OpenDover (http://opendover.nl/
), whose automatictasks are: classifying features of opinionated text by positioning them alonga continuum between sentiment polarities, such as negative/positive appraisal;extracting the components or attributes of the attitude in terms oflexicogrammatical token (noun, verb, adverb, adjective); labelling featuresaccording to the attitudes comprised in the Appraisal Framework (affect,judgment, appreciation). The Focused Exploration phase is sustained by thesetasks and confirms the first hypothesis. The Verification phase shows datatriangulation and the comparative analysis to confirm the second hypothesis.The author concludes that “sentiment analyzers open doors for advanced uses oftechniques already adopted in information extraction and text analysis indifferent disciplinary domains” (p. 168).
Chapter 8, ‘Reflections on Discursive Practices Mediated by Semantic Web Apps’(pp. 169-173), briefly summarizes the main issues of this work and offersinsights not only into how discursive practices on the Web produce meaningbut also “how the global context of the Web is providing more interactionalresources which enable a wider audience to craft new meaningful discursiveprocesses” (p. 171). Future work needs to look at language-related facetsregarding Web 3.0 and Web 4.0 phenomena such as multiple-authored texts,mashup applications, and intelligent personal agents.
“Applied Linguistics and Semantics Web Apps” is an authoritative and rigorousinquiry in two fields of study: Applied Linguistics and Semantic Web, twoterritories that in the age of the Web 2.0 - but 3.0 and 4.0 very shortly -necessitate a reciprocal osmosis since they are ancillary to each other. Mostdiscursive practices, in fact, occur on the Web and Semantic Web Apps can playa strategic role of ‘sharpening applied linguistics research’. This innovativehypothesis is widely discussed across the seven chapters, both theoreticallyand practically. The first part of the book shows an extensive knowledge ofthe major issues and theories as well as theoretical approaches to the tworesearch fields. The second part tests the hypotheses by putting into practicetheoretical concepts via three different but meaningful case studies.
In addition to being written in a clear and accessible style, the bookprovides the reader with all necessary tools, including resources availablefree online, data, tables, examples, screenshots, and finally an extensivebibliography to orient the reader within the enquiry. A specialised glossaryand/or indexes of name and subject at the end of the book would have added toits usability.
The book will appeal to scholars and researchers in Applied Linguistics,Psycholinguistics, Corpus Linguistics and Computational Linguistics. It canalso be used by language and content engineers as a theoretical and practicalsource thanks to which future innovations of web technology could be exploitedto improve Natural Language Processing systems. Furthermore, the book providesan excellent and much needed resource for post-graduate students and teacherswho already possess a well-developed digital literacy.
One of two slight limitations lies in this last claim. The scope of thisresearch enquiry and especially its practical applications require the readerto be digitally ‘literate’ (or “multiliterate” as Lemke 1998 puts it) in orderto use, apply and interpret Semantic Web Apps in professional, educational andsocial contexts. The author herself underlines this problem in Chapter 5 (p.120). It is therefore necessary to make a distinction between potentialaddressees of these apps whose aim is to speculate on language use andrepresentation on the Web (e.g., applied linguists) and those whose aim issimply to search for information on the Internet (e.g., patients or hotelconsumers). Conversely, the exploitation of Semantic Web Apps technology ineducational settings is more innovative and challenging for all participantsinvolved, i.e. students, teachers, applied linguists and, last but not least,software developers.
The second limitation is that a more detailed investigation on AffectiveComputing models in Chapter 7 would have been complementary to the sentimentdetection approach and would have offered a wider perspective as they havebeen widely influencing many areas of research, such as education, L2learning/teaching, medicine, professional training, marketing, entertainment,etc. (cf. Picard 2000; Picard et al. 2001, 2004; Wehrle 2000; Woolf et al.2009; Kopecek 2000).
To conclude, “Applied Linguistics and Semantics Web Apps” is noteworthy in itscombination of density and breadth of topics. Its novelty lies in having shownthe interesting challenge of integrating two different and apparently distantfields of research and exploring the potential for cross-permeability betweenApplied Linguistics and Semantic Web.
Aronson, E., Wilson, T.D., Akert, R.M. 2005. Social Psychology. 7th ed. UpperSaddle River, NJ: Person Education, Inc.
Faber, P., Marquez Linares, C., Vega Exposito, M. 2005. Framing Terminology. AProcess-Oriented Approach. Meta: journal des traductores/Meta: Translators'journal, 50, no. 4.http://www.erudit.org/revue/meta/2005/v50/n4/019916ar.pdf
Fillmore, C.J. 1982. Frame Semantics. The Linguistic Society of Korea (ed.).Linguistics in the Morning Calm. Seoul: Hanshin, pp. 111-137.
Fillmore, C.J. 1985. Frames and the Semantics of Understanding. Quaderni diSemantica, 6(2), 222-254.
Kopecek, I. 2000. Emotions and Prosody in Dialogues: An Algebraic ApproachBased on User Modelling. Proceedings of the ISCA Workshop on Speech andEmotions. Belfast: ISCA, pp. 184-189.
Lemke, J.L. 1998. Resources for Attitudinal Meaning: Evaluative orientationsin text semantics. Functions of Language 5(1), 33-56.
Martin, J.R. 2000. Beyond Exchange: APPRAISAL system in English. In S. Huston& G. Thompson (eds.). Evaluation in Text: Authorial Stance and theConstruction of Discourse. Oxford: Oxford University Press, pp. 142-175.
Martin, J.R., White, P.R.R. 2005. The Language of Evaluation: Appraisal inEnglish. London: Palgrave.
Paivio, A. 1986. Mental Representation: A dual-coding approach. New York:Cambridge University Press.
Paivio, A. 2006. Dual coding theory and education. Chapter prepared for thePathways to Literacy Achievement for High Poverty Children conference;University of Michigan, School of Education, Ann Arbor, September 29 October1, 2006, pp. 1-19.http://www.umich.edu/~rdytolrn/pathwaysconference/presentations/paivio.pdf
Picard, R.W. 1997. Affective Computing. Cambridge: The MIT Press.
Picard, R.W. 2000. An Interview with Rosalind Picard, Author of AffectiveComputing, in Affective Interactions. Towards a New Generation of ComputerInterfaces, A. Paiva (eds). Berlin: Springer, pp. 219-227.
Picard, R.W., Vyzas, E., Healey,J. 2001. Toward Machine EmotionalIntelligence: Analysis of Affective Physiological State. IEEE TransactionsPattern Analysis and Machine Intelligence, 23 (10), pp. 1175-1191.
Picard, R.W., Papert, S., Bender, W., Blumberg, B., Breazeal, C., Cavallo, D.,Machover, T., Resnick, M., Roy, D., Strohecker, C. 2004. Affective learning- a manifesto. BT Technical Journal, Vol. 2, No 4, pp. 253-269.
Wehrle, T. Kaiser. 2000. Emotion and Facial Expression, in AffectiveInteractions. Towards a New Generation of Computer Interfaces, A. Paiva (ed).Berlin: Springer, pp. 49-63.
Weidemann, A. 2003. Towards Accountability: A point of orientation forpost-modern applied linguistics in the third millennium. Literatur, 24(1),1-20.
Woolf, B., Burleson, W., Arroyo, I., Dragon, T., Cooper, D., Picard, R.W.2009. Affect-aware Tutors: recognising and responding to student affect.International Journal of Learning Technology 4(3/4). pp. 129-164.www1.cs.columbia.edu/~julia/papers/cowie00.pdf.
ABOUT THE REVIEWER
Sandra Petroni is a tenured researcher in English Language and Linguistics.She teaches at the Faculty of Humanities of the University of Rome “TorVergata”, on the Languages in the Information Society Degree Course. She isthe author of several research articles and two books, Self-Study. Lamultimedialità e l'apprendimento della lingua inglese nel nuovo sistemauniversitario italiano (2004) and Language in the Multimodal Web Domain(2011). She is also a member of national and international ScientificAssociations. Her research fields are: multimodality, specialized discourse -in particular the Information and Communication Technology (ICT) domain,cognitive semiotics and psycholinguistics.
Page Updated: 22-Jan-2013