Combining evidence with a probabilistic framework for answer ranking and answer merging in question answering

Question answering (QA) aims at finding exact answers to a user’s question from a large collection of documents. Most QA systems combine information retrieval with extraction techniques to identify a set of likely candidates and then utilize some ranking strategy to generate the final answers. This ranking process can be challenging, as it entails identifying the relevant answers amongst many irrelevant ones. This is more challenging in multi-strategy QA, in which multiple answering agents are used to extract answer candidates. As answer candidates come from different agents with different score distributions, how to merge answer candidates plays an important role in answer ranking. In this paper, we propose a unified probabilistic framework which combines multiple evidence to address challenges in answer ranking and answer merging. The hypotheses of the paper are that: (1) the framework effectively combines multiple evidence for identifying answer relevance and their correlation in answer ranking, (2) the framework supports answer merging on answer candidates returned by multiple extraction techniques, (3) the framework can support list questions as well as factoid questions, (4) the framework can be easily applied to a different QA system, and (5) the framework significantly improves performance of a QA system. An extensive set of experiments was done to support our hypotheses and demonstrate the effectiveness of the framework. All of the work substantially extends the preliminary research in Ko et al. (2007a). A probabilistic framework for answer selection in question answering. In: Proceedings of NAACL/HLT.     

Investigating social networks of older Singaporean learners: a mixed methods approach

ABSTRACT

Lifelong learning is regarded as an important channel in promoting active engagement in later life for aging societies. While most studies depict older learners as a group resilient to engaging in lifelong learning, few have addressed the impacts of their social networks on their participation. Drawing on the nationwide lifelong learning program in Singapore, the study explores the extent older Singaporean adults’ social networks influence their involvement in learning courses and illustrates how those networks matter to their motivations. A mixed methods approach consisting of two network instruments (Name Generator and Position Generator) and in-depth interviews based on 30 older Singaporeans (between 50 and 79 years old) were employed. The findings demonstrated that primary family members (spouse and children) were key discussants for older learners, but the narratives showed only children were key supporters. For female learners, husbands’ support could be limited. Such a gender difference was revealed in overlap networks among couples, with male learners receiving greater positive support from their wives. Furthermore, the results showed that single or widowed learners had more non-kin members and diverse network resources, which reflect in their discourses of being highly motivated and active in spreading news of courses. To conclude, the study delivered deeper understandings of how diverse social contexts influence older learners’ motivations. Future research shall continue to focus on variations of network characteristics and network resources to improve the understanding of how significant others and accessible network resources provide social support or opportunities for older learners.     

Propolis modulates vitronectin, laminin, and heparan sulfate/heparin expression during experimental burn healing

Objective

This study was aimed at assessing the dynamics of vitronectin (VN), laminin (LN), and heparan sulfate/heparin (HS/HP) content changes during experimental burn healing.

Methods

VN, LN, and HS/HP were isolated and purified from normal and injured skin of domestic pigs, on the 3rd, 5th, 10th, 15th, and 21st days following thermal damage. The wounds were treated with apitherapeutic agent (propolis), silver sulfadiazine (SSD), physiological salt solution, and propolis vehicle. VN and LN were quantified using an immunoenzymatic assay and HS/HP was estimated by densitometric analysis.

Results

Propolis treatment stimulated significant increases in VN, LN, and HS/HP contents during the initial phase of study, followed by a reduction in the estimated extracellular matrix molecules. Similar patterns, although less extreme, were observed after treatment with SSD.

Conclusions

The beneficial effects of propolis on experimental wounds make it a potential apitherapeutic agent in topical burn management.     

The Problem of Pseudoscience in Science Education and Implications of Constructivist Pedagogy

The intrusion of pseudoscience into science classrooms is a problem in science education today. This paper discusses the implications of constructivist pedagogy, which relies on the notions of viability and inter-subjectivity, in a context favourable to the acceptance of pseudoscience. Examples from written statements illustrate how prospective science teachers in Turkey readily accept pseudoscientific explanations of the origin of species. Constructivist pedagogy underestimates, if not ignores, the difficulty of holding rational discussions in the presence of pseudoscientific or absolute beliefs. Moreover, it gives a higher priority to learners’ exposure to alternative constructions through social negotiation than to furthering their appreciation of science. Under these circumstances, self-confirmation and social pressure to accept existing pseudoscientific beliefs may be unanticipated consequences of social negotiation. Considering the aim of science education to foster an appreciation of science, the implications of constructivist pedagogy are, or should be, of great concern to science educators.     

Argumentation in elementary science education: addressing methodological issues and conceptual understanding

In this review essay I respond to issues raised in Mijung Kim and Wolff-Michael Roth’s paper titled “Dialogical argumentation in elementary science classrooms”, which presents a study dealing with dialogical argumentation in early elementary school classrooms. Since there is very limited research on lower primary school students’ argumentation in school science, their paper makes a contribution to research on children’s argumentation skills. In this response, I focus on two main issues to extend the discussion in Kim and Roth’s paper: (a) methodological issues including conducting a quantitative study on children’s argumentation levels and focusing on children’s written argumentation in addition to their dialogical argumentation, and (b) investigating children’s conceptual understanding along with their argumentation levels. Kim and Roth emphasize the difficulty in determining the level of children’s argumentation through the Toulmin’s Argument Pattern and lack of high level arguments by children due to their difficulties in writing texts. Regarding these methodological issues, I suggest designing quantitative research on coding children’s argument levels because such research could potentially provide important findings on children’s argumentation. Furthermore, I discuss alternative written products including posters, figures, or pictures generated by children in order to trace children’s arguments, and finally articulating argumentation and conceptual understanding of children.     

From FRA to RFN,or How the Family Resemblance Approach Can Be Transformed for Science Curriculum Analysis on Nature of Science

The inclusion of Nature of Science (NOS) in the science curriculum has been advocated around the world for several decades. One way of defining NOS is related to the family resemblance approach (FRA). The family resemblance idea was originally described by Wittgenstein. Subsequently, philosophers and educators have applied Wittgenstein’s idea to problems of their own disciplines. For example, Irzik and Nola adapted Wittgenstein’s generic definition of the family resemblance idea to NOS, while Erduran and Dagher reconceptualized Irzik and Nola’s FRA-to-NOS by synthesizing educational applications by drawing on perspectives from science education research. In this article, we use the terminology of “Reconceptualized FRA-to-NOS (RFN)” to refer to Erduran and Dagher’s FRA version which offers an educational account inclusive of knowledge about pedagogical, instructional, curricular and assessment issues in science education. Our motivation for making this distinction is rooted in the need to clarify the various accounts of the family resemblance idea.The key components of the RFN include the aims and values of science, methods and methodological rules, scientific practices, scientific knowledge as well as the social-institutional dimensions of science including the social ethos, certification, and power relations. We investigate the potential of RFN in facilitating curriculum analysis and in determining the gaps related to NOS in the curriculum. We analyze two Turkish science curricula published 7 years apart and illustrate how RFN can contribute not only to the analysis of science curriculum itself but also to trends in science curriculum development. Furthermore, we present an analysis of documents from USA and Ireland and contrast them to the Turkish curricula thereby illustrating some trends in the coverage of RFN categories. The results indicate that while both Turkish curricula contain statements that identify science as a cognitive-epistemic system, they underemphasize science as a social-institutional system. The comparison analysis shows results such as the “scientific ethos” category being mentioned by the Irish curriculum while “social organizations and interactions” category being mentioned by the Turkish curriculum. In all documents, there was no overall coherence to NOS as a holistic narrative that would be inclusive of the various RFN categories simultaneously. The article contributes to the framing of NOS from a family resemblance perspective and highlights how RFN categories can be used as analytical tools.     

Parametrically excited oscillation of stay cable and its control in cable-stayed bridges

This paper presents a nonlinear dynamic model for simulation and analysis of a kind of parametrically excited vibration of stay cable caused by support motion in cable-stayed bridges. The sag, inclination angle of the stay cable are considered in the model, based on which, the oscillation mechanism and dynamic response characteristics of this kind of vibration are analyzed through numerical, calculation. It is noted that parametrically excited oscillation of a stay cable with certain, sag, inclination angle and initial static tension force may occur in cable-stayed bridges due to deck vibration under the condition that the natural frequency of a cable approaches to about half of the, first model frequency of the bridge deck system. A new vibration control system installed on the cable anchorage is proposed as a possible damping system to suppress, the cable parametric oscillation. The numerical calculation results showed that with the use of this damping system, the cable oscillation due to the vibration of the deck and/or towers will be considerably reduced. Project supported by National Natural Science Foundation of China (No. 59978044)     

Pedagogy and human dignity – the Special Rank Teacher in China since 1978

Teachers in China are regarded as vulnerable to sociopolitical changes. This paper, however, focuses on the resilience and innovativeness of state‐selected expert teachers – the recipients of the Special Rank Teacher (SRT) award. This award was the product of a transitional period in the aftermath of the Cultural Revolution, and was an act of rehabilitation for the teaching profession. Using a biographical approach, this paper analyses how the intentions of the award were reflected in the choice of recipients. It contributes to a deeper understanding of teachers in China and the notion of teaching excellence that the state‐recognised teachers exemplify.     

The education for sustainable development movement in Japan: a political perspective

The Japanese government provided various political opportunities for non‐governmental groups and individuals in Japan to ‘jointly propose’ policy on education and sustainable development at the World Summit on Sustainable Development, Johannesburg, 2002. These opportunities resulted in the emergence of the Japanese education for sustainable development (ESD) movement, and the crystallisation of a broader proposal that led to the initiation of the UN Decade of ESD (2005–2014). In this paper, we trace the history of these two outcomes, arguing that the opportunities, developed through the coordination of non‐governmental groups by government, took place within, rather than broadened or confronted, the government’s scope of interests. While the paper illustrates how the government’s continued support was crucial to the development of the ESD movement and the UN Decade, and the movement has met with considerable achievements thus far (via its collective challenges to conventional education in a sustainability context in Japan), we argue that recognition of the political opportunity structures that affect the movement’s further development remains crucial. In particular, we argue for close attention to the significance of a corporatist framing of this emerging civil society movement in Japan by the national government, and call for further political and historical analysis of ESD movements and their relations with government, around the world.