P.A.J.M. Coppen

Evaluating discourse-based answer extraction for why -question answering

30th Annual International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR 2007)

What is not in the bag of words for why-qa?

While developing an approach to why-QA, we extended a passage retrieval system that uses off-the-shelf retrieval technology with a re-ranking step incorporating structural information. We get significantly higher scores in terms of MRR150 (from 0.25 to 0.34) and success10. The 23 improvement that we reach in terms of MRR is comparable to the improvement reached on different QA tasks by other researchers in the field, although our re-ranking approach is based on relatively lightweight overlap measures incorporating syntactic constituents, cue words, and document structure.

Defining Grammatical Difficulty: A Student Teacher Perspective.

Numerous second language acquisition (SLA) researchers have tried to define grammatical difficulty in second and foreign language acquisition – often as part of an attempt to relate the efficacy of different types of instruction to the degree of difficulty of grammatical structures. The resulting proliferation of definitions and the lack of a unifying framework have made the easy–difficult distinction a muddled concept. This paper attempts to clear up some of the confusion by reviewing the definitions SLA literature offers and relating these to (student) teachers’ cognitions on grammatical difficulty, a perspective largely ignored in SLA publications despite its potential for a more integrative and holistic view on the topic. To this end, a total of 727 undergraduate and postgraduate student teachers of English were surveyed in two empirical studies. Quantitative and qualitative analyses – in combination with the findings from SLA literature – yielded a model comprising four interrelated categories of factors that determine grammatical difficulty: (1) grammar feature; (2) pedagogical arrangement; (3) teacher quality; and (4) learner characteristics. In addition, the resulting model is discussed as a basis for future research into the changing cognitions of student teachers as they become more experienced.

Student teacher beliefs on grammar instruction

The role of grammar teaching in foreign language education is a controversial one both in second language acquisition (SLA) research and language pedagogy and, as a result, a potential source of confusion to student teachers. The objective of this study was to gain insight into the beliefs on grammar teaching of student teachers of English as a foreign language enrolled in undergraduate and postgraduate teacher education programmes at Dutch universities of applied sciences. To this end a questionnaire was developed and validated based on four construct pairs from SLA literature: meaning- versus form-focused instruction, focus on form (FonF) versus focus on forms (FonFs), implicit versus explicit instruction, and inductive versus deductive instruction. Overall, respondents (n = 832) were found to prefer form-focused, explicit, inductive instruction, and FonFs. However, higher-year undergraduates’ and postgraduates’ results showed a trend towards a preference for more meaning-focused and implicit instruction, and FonF. When learner level was factored in, however, these forms of language instruction were considered subordinate to more traditional form-focused approaches for teaching higher-level language learners.

Evaluating paragraph retrieval for why-QA

We implemented a baseline approach to why-question answering based on paragraph retrieval. Our implementation incorporates the QAP ranking algorithm with addition of a number of surface features (cue words and XML markup). With this baseline system, we obtain an accuracy-at-10 of 57.0% with an MRR of 0.31. Both the baseline and the proposed evaluation method are good starting points for the current research and for other researchers working on the problem of why-QA. We also experimented with the addition of smart question analysis features to our baseline system (answer type and informational value of the subject). This however did not give significant improvement to our baseline. In the near future, we will investigate what other linguistic features can facilitate re-ranking in order to increase accuracy.

Bridging the gap between linguistic theory and L1 grammar education - experts' views on essential linguistic concepts

ABSTRACT L1 grammar education is internationally criticised because of its pedagogy and its curriculum content. There is a gap between linguistic theory and school grammar in which the latter rarely makes use of possibly relevant insights from the former. At the same time, linguistics itself has never seriously undertaken attempts to identify the fundamental, theory-neutral conceptual knowledge of the field. Instead, most linguists spend time on defining the boundaries between different linguistic theories, making it harder for language teachers to take advantage of insights that linguistics has to offer. This paper, therefore, aims (1) to establish a theory-neutral foundation for sentential syntax and semantics and (2) to explore the role this foundation should play in language education according to linguists. The experts were asked to articulate the most important linguistic concepts for both the domain of linguistic theory and the domain of grammar education. Twenty-six concepts were identified and ranked in a Delphi study for relative importance in both domains. The importance of the concepts correlated strongly for both domains, making it feasible to bridge the gap between linguistics and school grammar. The Dutch context is taken as a frame of reference, although the studys relevance is likely to be broader.

Linguistic concepts in L1 grammar education: a systematic literature review

ABSTRACT Teaching grammar has always constituted a major part of language education in curricula around the world, although it has also been heavily debated. Most of the debate on grammar teaching focused on the rationales for teaching it, rather than on the linguistic content that should be taught. At the same time, there appears to be a renewed interest in restoring the bond between linguistic theory and grammar education. Previous research has suggested that it would be highly desirable to gain a clearer picture of this content. Which concepts are being discussed in the literature on grammar education, and to what extent are these compatible with modern linguistics, in other words: is the literature on grammar teaching up-to-date? This systematic literature review is the first to dive into these questions. Results indicate: (1) most of the concepts in the literature on grammar teaching are from traditional grammar. To a limited extent, there are also concepts from modern linguistic theory that are being discussed, but mostly implicitly; (2) most concepts are not being motivated because they are meaningful in modern theoretical linguistics, but because they reflect traditional classroom practices and policy. It can consequently be concluded that education on linguistic analysis is not up-to-date, which potentially has severe consequences: implementing insights from modern linguistics is likely to provide students with deeper insights, and teachers with a better equipped pedagogy.

RLIST: a representation scheme for the automated comparison of psychopathological scales and syndromes

Abstract Many different instruments for the assessment and diagnosis of psychopathology are used in research and clinical practice. A representation scheme, ‘RLIST’, was developed for examining the relations between the syndromes or syndrome scales of different instruments by means of the computer. This representation scheme allows the descriptive similarity of syndromes to be determined with more precision than by the manual or approximate approaches presented in the literature. The focus of the present study was a detailed discussion of the syntax and semantics of RLIST, using a comparison of DSM and CBCL classified syndromes as an example. Although originally developed to compare these syndromes, RLIST can be used to match syndromes of other instruments as well. RLIST is capable of representing various kinds of syndrome definitions, including syndromes that require all symptoms to be present (monothetic) and syndromes that require only an optional and minimal subset of symptoms to be present (polythetic). In addition to expressing the logical operators AND, OR, and NOT, RLIST is able to express other formulas of propositional logic, namely the exclusive OR, Equivalence, and Implication. Compared with set theory and established representation schemes such as production rules, RLIST has several advantages. For instance, RLIST is highly compact and permits the weighting of symptoms.