Aggregation of preferences in recommender systems

This chapter gives an overview of aggregation functions toward their use in recommender systems. Simple aggregation functions such as the arithmetic mean are often employed to aggregate user features, item ratings, measures of similarity, etc., however many other aggregation functions exist which could deliver increased accuracy and flexibility to many systems. We provide definitions of some important families and properties, sophisticated methods of construction, and various examples of aggregation functions in the domain of recommender systems.

Student preferences for bachelor degrees at TAFE: the socio-spatial influence of schools

This report on Student Preferences for Bachelor Degrees at TAFE (Technical and Further Education) institutions is derived from research commissioned by Australia’s National Centre for Student Equity in Higher Education (NCSEHE) hosted at Curtin University and conducted by researchers at Deakin University’s Strategic Centre for Research in Educational Futures and Innovation (CREFI). The report focuses on the influence of schools on their students’ higher education (HE) preferences – particularly their preferences for TAFE bachelor degrees – as recorded by the Victorian and South Australian Tertiary Admissions Centres (VTAC and SATAC). Influence is researched in terms of a school’s socioeconomic status, geographical location and sector. The SATAC data set is considerably smaller, at around 8 per cent of the VTAC data set.Bachelor degrees offered by TAFEs are relatively small in number but a growing higher education option for students in Australia (Gale et al. 2013). The Australian Government’s proposal to extend Commonwealth Supported Places (CSPs) to include Australian higher education not delivered by the nation’s public universities (Department of Education 2014b), is likely to fuel further growth in TAFE bachelor degree offerings. The recent Report of the Review of the Demand Driven Funding System in Australian higher education (Kemp & Norton 2014), which recommended this change, also makes special mention of non-university degree options as something that would be of particular benefit to students from low socioeconomic status backgrounds.The research reported herein is informed by a review of the international research literature, which indicates three main influences on students’ HE preferences: (1) students’ families and communities; (2) the socio-spatial location of their schools; and (3) school practices. This report contributes to understandings on the second of these: the influence of school context (their socio-spatial location) on students’ preferences for TAFE bachelor degrees.The research found that the annual rate of student preferences for TAFE bachelor degrees was relatively stable (at around 1,500 per annum) from 2009 to 2012 but rose significantly (by 30%) in 2013. Students from high socioeconomic status schools (and with an average ATAR of 56.9) were the group that registered the largest number of preferences. The number of preferences for TAFE bachelor degrees lodged by students from metropolitan schools exceeded the preferences of students from schools located in all other regions combined. This might reflect the fact that TAFE institutions offering bachelor degrees tend to be located in metropolitan areas.The research also found that students’ preferences for TAFE bachelor degrees increased after announcement of their Australian Tertiary Admission Rank (ATAR), by between 25 and 30 per cent each year. The post-ATAR increase was most noticeable in the Health and Education fields of study and among students from high socioeconomic status schools. The report concludes that while the public perception of TAFE is that it is a sector primarily for students from low SES backgrounds, this is not reflected in students’ preferences for TAFE bachelor degrees. Instead, the preferences of students from high socioeconomic schools outnumber other SES groups in almost every TAFE-degree field of study. This includes the fields of Health and Education, which are often seen to be typical low SES student choices in universities (Gale & Parker 2013).

Optimal redistribution with hetrogeneous preferences for leisure

This paper examines the properties of the optimal nonlinear income tax when preferences are quasi–linear in leisure and individuals differ in their ability and their preferences for leisure. The government seeks to redistribute income. It can perfectly observe the level of endogenous income but cannot observe either ability or preferences. The heterogeneity of preferences leads to problems of comparability between individual utilities which challenge the design of redistributive schemes. We analyze the consequences of adopting a utilitarian social welfare function where the government is allowed to give different weights to individuals with different preferences. Under this particular social objective and given the quasi–linearity of preferences, we are able to obtain closed–form solutions for the marginal tax rates and to examine the progressivity of the tax system according to the weights used.

Migration, unemployment and trade

Migration, Unemployment and Trade focuses on the issues of migration, welfare and unemployment in a trade and development framework. Several chapters of the book analyze the implications of internal labor mobility in a model designed to highlight its implications for regional welfare, urban unemployment, rural-urban dichotomy and structural adjustment. An important innovation in this work is the disaggregation of the economy and the use of separate utility functions to highlight non-homogeneity of preferences. The book also deals with international mobility of factors in different frameworks. In particular it concentrates on the highly emotive issue of legal and illegal migration. Thus this work incorporates interesting and important features of labor economics and factor mobility into trade and distortion theory.

Teacher subject specialisms and their relationships to learning styles, psychological types and multiple intelligences: implications for course development

This study explores issues in teacher education that increase our understanding of, and response to, the individual differences displayed by learners. A large undergraduate teacher education cohort provided evidence of the range and distribution of preferences in learning styles, psychological types and multiple intelligences. This information revealed that distributions of scores on the Kolb Learning Style Inventory, the Myers-Briggs Type Indicator, and the Multiple Intelligences Checklist for Adults provide evidence about the scope and range of differences between four teacher subject specialisms. This rich information about those participating in teacher education courses provides some guidance for educating those with their own clear preferences to the range of different preferences expressed by many other learners and highlights the existence of four sets of major differences in approaches to teaching and learning in prospective teachers.

Modelling regular and estimable inverse demand systems : a distance function approach

To be useful for policy simulation in the current climate of rapid structural change, inverse demand systems must remain regular over substanstial variations in quantities. The distance function is a convenient vehicle for generating such systems. It also allows convenient imposition of prior ideas about the structure of preferences required for realistic policy work. While the distance function directly yields Hicksian inverse demand functions via the Shepard-Hanoch lemma, they are usually explicit in the unobservable level of utility (u), but lack a closed-form representation in terms of the observable variables. Note however that the unobservability of u need not hinder estimation. A simple one-dimensional numerical inversion allows the estimation of the distance function via the parameters of the implied Marshallian inverse demand functions. This paper develops the formal theory for using distance functions in this context, and reports on initial trials on the operational feasibility of the method.

A two-stage dynamic group decision making method for processing ordinal information

In group decision making (GDM) problems, ordinal data provide a convenient way of articulating preferences from decision makers (DMs). A number of GDM models have been proposed to aggregate such kind of preferences in the literature. However, most of the GDM models that handle ordinal preferences suffer from two drawbacks: (1) it is difficult for the GDM models to manage conflicting opinions, especially with a large number of DMs; and (2) the relationships between the preferences provided by the DMs are neglected, and all DMs are assumed to be of equal importance, therefore causing the aggregated collective preference not an ideal representative of the group's decision. In order to overcome these problems, a two-stage dynamic group decision making method for aggregating ordinal preferences is proposed in this paper. The method consists of two main processes: (i) a data cleansing process, which aims to reduce the influence of conflicting opinions pertaining to the collective decision prior to the aggregation process; as such an effective solution for undertaking large-scale GDM problems is formulated; and (ii) a support degree oriented consensus-reaching process, where the collective preference is aggregated by using the Power Average (PA) operator; as such, the relationships of the arguments being aggregated are taken into consideration (i.e., allowing the values being aggregated to support each other). A new support function for the PA operator to deal with ordinal information is defined based on the dominance-based rough set approach. The proposed GDM model is compared with the models presented by Herrera-Viedma et al. An application related to controlling the degradation of the hydrographic basin of a river in Brazil is evaluated. The results demonstrate the usefulness of the proposed method in handling GDM problems with ordinal information.

Learning preferences of TAFE and university students

Using Entwistle and Ramsden's (1983) Approaches to Studying Inventory, Fuller and Chalmers showed little difference between TAFE students and university students both in the factor structures derived through factor analysis, and in a comparison of individual subscale through analysis of variance. Fuller and Chalmers had hypothesised that differences in approaches to study may be identified as a result of the different learning and teaching contexts experienced by the two groups of students. The current study has examined the similarities and differences between students in the two sectors using a learning preferences inventory. Learning preferences were chosen for investigation on the basis of previous theoretical work which suggests that it is preferences that are most influenced by environmental factors. The present study has shown significant similarity between the learning preferences of the two groups of students. However, there were also significant differences which indicate that the Fuller and Chalmers hypothesised differences may be more observable at a preferences level than at an approaches to study level.

Technology student learning preferences and the design of flexible learning programs

The learning preferences of three hundred and thirty eight technology students enrolled in sub-degree programs at an Australian institution of Technical and Further Education were tested using the Canfield Learning Styles Inventory (CLSI). The results have been interpreted in a learning preferences framework and provide supportive evidence for the preferences factors of print-nonprint, collaborative, dependent,and autonomous learning identified by Sadler-Smith & Riding (1999). Although there search focussed on learning preferences the analysis also indicated support for the Wholist-Analytic cognitive style proposed by Riding & Cheema (1991). Gender differences were shown for the Interest subscales of the CLSI. Age-group differences were shown for several Conditions of Learning and Modes of Learning subscales. Implications for the design of training programs, and the skills that may need to be developed in technology learners to enable effective use of flexible delivery, are also discussed.

Refugees and residents: densities and habitat preferences of lorikeets in urban Melbourne

Lorikeet densities were measured across four habitat types in urban Melbourne. Musk Glossopsitta concinna and Rainbow Lorikeets Trichoglossus haematodus were shown to preferentially use established streetscapes with predominantly native vegetation. The high densities of Musk Lorikeets recorded possibly reflect a paucity of flowering in Victorian BoxIronbark forests during the autumnlwinter of 2002 and the availability of supplementary nectar resources in the urban environment. Future planting decisions in recently developed streetscapes will dictate the long-term resource potential for lorikeets and other nectarivores in urban Melbourne.

Managing the family`s health: preferences and experiences of women with different work and family arrangements.

Balancing the needs of work and family is a subject of much debate.The purpose of this research was to explore how families manage their children's health within the context of different work and family arrangements.In-depth interviews were conducted with women who were at home full time (8) or in paid work over 30 hours a week (7). Women had at least one child under the five years of age. Findings revealed there was no simple relationship between women's working arrangements and how they managed their children's health. All women, irrespective of their working arrangements, held similar preferences for managing their children's health.However, most women experienced either time or financial constraints that meant they had to compromise their original preferences. In some cases this meant children missed out on receiving health services. Workplace support, extended family support and general satisfaction with work and family arrangements appeared to be important factors for the small number of women who had no problems in managing their children's health. The implications of these findings are discussed.

The effects of television advertisements for junk food versus nutritious food on children's food attitudes and preferences

Television (TV) food advertising has attracted criticism for its potential role in promoting unhealthy dietary practices among children. Content analyses indicate junk food advertising is prevalent on Australian children's TV; healthy eating is rarely promoted. This paper presents (a) a cross-sectional survey examining associations between children's regular TV viewing habits and their food-related attitudes and behaviour; and (b) an experiment assessing the impact of varying combinations of TV advertisements (ads) for unhealthy and healthy foods on children's dietary knowledge, attitudes and intentions. The experimental conditions simulated possible models for regulating food ads on children's TV. Participants were 919 grade five and six students from schools in Melbourne, Australia. The survey showed that heavier TV use and more frequent commercial TV viewing were independently associated with more positive attitudes toward junk food; heavier TV use was also independently associated with higher reported junk food consumption. The experiment found that ads for nutritious foods promote selected positive attitudes and beliefs concerning these foods. Findings are discussed in light of methodological issues in media effects research and their implications for policy and practice. It is concluded that changing the food advertising environment on children's TV to one where nutritious foods are promoted and junk foods are relatively unrepresented would help to normalize and reinforce healthy eating.

Choosing ICT? A segmentation analysis of students' preferences for hybrid study mode

Students’ acceptance and use of ICT-based learning needs to be understood in terms of their preferences for alternatives such as face-to-face (FtF) or print-based learning. This paper reports on an investigation of students’ preferences for hybrid study modes. Cluster analysis was used to identify segments of students that had distinctive preferences for combinations of FtF, print and web-based study modes. Five segments were identified. These segments were distinguishable on some demographic and situational characteristics. The size and nature of the segments have implications for the hybrid modes offered by universities and the extent to which students’ may embrace the ICT-based innovations designed by educators.

Cancer patient preferences for communication of prognosis in the metastatic setting

PURPOSE: To identify preferences for and predictors of prognostic information among patients with incurable metastatic cancer. PATIENTS AND METHODS: One hundred twenty-six metastatic cancer patients seeing 30 oncologists at 12 outpatient clinics in New South Wales, Australia, participated in the study. Patients were diagnosed with incurable metastatic disease within 6 weeks to 6 months of recruitment. Patients completed a survey eliciting their preferences for prognostic information, including type, quantity, mode, and timing of presentation; anxiety and depression levels; and information and involvement preferences. RESULTS: More than 95% of patients wanted information about side effects, symptoms, and treatment options. The majority wanted to know longest survival time with treatment (85%), 5-year survival rates (80%), and average survival (81%). Words and numbers were preferred over pie charts or graphs. Fifty-nine percent (59%) wanted to discuss expected survival when first diagnosed with metastatic disease. Thirty-eight percent and 44% wanted to negotiate when expected survival and dying, respectively, were discussed. Patients with higher depression scores were more likely to want to know shortest time to live without treatment (P = .047) and average survival (P = .049). Lower depression levels were significantly associated with never wanting to discuss expected survival (P = .03). Patients with an expected survival of years were more likely to want to discuss life expectancy when first diagnosed with metastases (P = .02). CONCLUSION: Most metastatic cancer patients want detailed prognostic information but prefer to negotiate the extent, format, and timing of the information they receive from their oncologists.