Term associations in query expansion : a structural linguistic perspective

Many successful query expansion techniques ignore information about the term dependencies that exist within natural language. However, researchers have recently demonstrated that consistent and significant improvements in retrieval effectiveness can be achieved by explicitly modelling term dependencies within the query expansion process. This has created an increased interest in dependency-based models. State-of-the-art dependency-based approaches primarily model term associations known within structural linguistics as syntagmatic associations, which are formed when terms co-occur together more often than by chance. However, structural linguistics proposes that the meaning of a word is also dependent on its paradigmatic associations, which are formed between words that can substitute for each other without effecting the acceptability of a sentence. Given the reliance on word meanings when a user formulates their query, our approach takes the novel step of modelling both syntagmatic and paradigmatic associations within the query expansion process based on the (pseudo) relevant documents returned in web search. The results demonstrate that this approach can provide significant improvements in web re- trieval effectiveness when compared to a strong benchmark retrieval system.

Using internet search queries for infectious disease surveillance: screening diseases for suitability

Background Internet-based surveillance systems provide a novel approach to monitoring infectious diseases. Surveillance systems built on internet data are economically, logistically and epidemiologically appealing and have shown significant promise. The potential for these systems has increased with increased internet availability and shifts in health-related information seeking behaviour. This approach to monitoring infectious diseases has, however, only been applied to single or small groups of select diseases. This study aims to systematically investigate the potential for developing surveillance and early warning systems using internet search data, for a wide range of infectious diseases. Methods Official notifications for 64 infectious diseases in Australia were downloaded and correlated with frequencies for 164 internet search terms for the period 2009–13 using Spearman’s rank correlations. Time series cross correlations were performed to assess the potential for search terms to be used in construction of early warning systems. Results Notifications for 17 infectious diseases (26.6%) were found to be significantly correlated with a selected search term. The use of internet metrics as a means of surveillance has not previously been described for 12 (70.6%) of these diseases. The majority of diseases identified were vaccine-preventable, vector-borne or sexually transmissible; cross correlations, however, indicated that vector-borne and vaccine preventable diseases are best suited for development of early warning systems. Conclusions The findings of this study suggest that internet-based surveillance systems have broader applicability to monitoring infectious diseases than has previously been recognised. Furthermore, internet-based surveillance systems have a potential role in forecasting emerging infectious disease events, especially for vaccine-preventable and vector-borne diseases

Improving spoken term detection using complementary information

This research has made contributions to the area of spoken term detection (STD), defined as the process of finding all occurrences of a specified search term in a large collection of speech segments. The use of visual information in the form of lip movements of the speaker in addition to audio and the use of topic of the speech segments, and the expected frequency of words in the target speech domain, are proposed. By using these complementary information, improvement in the performance of STD has been achieved which enables efficient search of key words in large collection of multimedia documents.

Query-by-Example Spoken Term Detection ALBAYZIN 2012 evaluation: overview, systems, results, and discussion

Query-by-Example Spoken Term Detection (QbE STD) aims at retrieving data from a speech data repository given an acoustic query containing the term of interest as input. Nowadays, it has been receiving much interest due to the high volume of information stored in audio or audiovisual format. QbE STD differs from automatic speech recognition (ASR) and keyword spotting (KWS)/spoken term detection (STD) since ASR is interested in all the terms/words that appear in the speech signal and KWS/STD relies on a textual transcription of the search term to retrieve the speech data. This paper presents the systems submitted to the ALBAYZIN 2012 QbE STD evaluation held as a part of ALBAYZIN 2012 evaluation campaign within the context of the IberSPEECH 2012 Conference(a). The evaluation consists of retrieving the speech files that contain the input queries, indicating their start and end timestamps within the appropriate speech file. Evaluation is conducted on a Spanish spontaneous speech database containing a set of talks from MAVIR workshops(b), which amount at about 7 h of speech in total. We present the database metric systems submitted along with all results and some discussion. Four different research groups took part in the evaluation. Evaluation results show the difficulty of this task and the limited performance indicates there is still a lot of room for improvement. The best result is achieved by a dynamic time warping-based search over Gaussian posteriorgrams/posterior phoneme probabilities. This paper also compares the systems aiming at establishing the best technique dealing with that difficult task and looking for defining promising directions for this relatively novel task.

Contribution à l'étude de la construction des concepts scientifiques au cours de l'apprentissage par problèmes en médecine

L’approche d’apprentissage par problèmes (APP) a vu le jour, dans sa forme contemporaine, à la Faculté de médecine de l’Université MacMaster en Ontario (Canada) à la fin des années 1960. Très rapidement cette nouvelle approche pédagogique active, centrée sur l’étudiant et basée sur les problèmes biomédicaux, va être adoptée par de nombreuses facultés de médecine de par le monde et gagner d’autres disciplines. Cependant, malgré ce succès apparent, l’APP est aussi une approche controversée, notamment en éducation médicale, où elle a été accusée de favoriser un apprentissage superficiel. Par ailleurs, les étudiants formés par cette approche réussiraient moins bien que les autres aux tests évaluant l’acquisition des concepts scientifiques de base, et il n’a jamais été prouvé que les médecins formés par l’APP seraient meilleurs que les autres. Pour mieux comprendre ces résultats, la présente recherche a voulu explorer l’apprentissage de ces concepts scientifiques, en tant que processus de construction, chez des étudiants formés par l’APP, à la Faculté de médecine de l’Université de Montréal, en nous appuyant sur le cadre théorique socioconstructivisme de Vygotski. Pour cet auteur, la formation des concepts est un processus complexe de construction de sens, en plusieurs étapes, qui ne peut se concevoir que dans le cadre d’une résolution de problèmes. Nous avons réalisé une étude de cas, multicas, intrasite, les cas étant deux groupes de neuf étudiants en médecine avec leur tuteur, que nous avons suivi pendant une session complète de la mi-novembre à la mi-décembre 2007. Deux grands objectifs étaient poursuivis: premièrement, fournir des analyses détaillées et des matériaux réflectifs et théoriques susceptibles de rendre compte du phénomène de construction des concepts scientifiques de base par des étudiants en médecine dans le contexte de l’APP. Deuxièmement, explorer, les approches de travail personnel des étudiants, lors de la phase de travail individuel, afin de répondre à la question de recherche suivante : Comment la dynamique pédagogique de l’APP en médecine permet-elle de rendre compte de l’apprentissage des concepts scientifiques de base? Il s’agissait d’une étude qualitative et les données ont été recueillies par différents moyens : observation non participante et enregistrement vidéo des tutoriaux d’APP, interview semi-structuré des étudiants, discussion avec les tuteurs et consultation de leurs manuels, puis traitées par diverses opérations: transcription des enregistrements, regroupement, classification. L’analyse a porté sur des collections de verbatim issus des transcriptions, sur le suivi de la construction des concepts à travers le temps et les sessions, sur le role du tuteur pour aider au développement de ces concepts Les analyses suggèrent que l’approche d’APP est, en général, bien accueillie, et les débats sont soutenus, avec en moyenne entre trois et quatre échanges par minute. Par rapport au premier objectif, nous avons effectivement fourni des explications détaillées sur la dynamique de construction des concepts qui s'étend lors des trois phases de l'APP, à savoir la phase aller, la phase de recherche individuelle et la phase retour. Pour chaque cas étudié, nous avons mis en évidence les représentations conceptuelles initiales à la phase aller, co-constructions des étudiants, sous la guidance du tuteur et nous avons suivi la transformation de ces concepts spontanés naïfs, lors des discussions de la phase retour. Le choix du cadre théorique socio constructiviste de Vygotski nous a permis de réfléchir sur le rôle de médiation joué par les composantes du système interactif de l'APP, que nous avons considéré comme une zone proximale de développement (ZPD) au sens élargi, qui sont le problème, le tuteur, l'étudiant et ses pairs, les ressources, notamment l'artefact graphique carte conceptuelle utilisée de façon intensive lors des tutoriaux aller et retour, pour arriver à la construction des concepts scientifiques. Notre recherche a montré qu'en revenant de leurs recherches, les étudiants avaient trois genres de représentations conceptuelles: des concepts corrects, des concepts incomplets et des concepts erronés. Il faut donc que les concepts scientifiques théoriques soient à leur tour confrontés au problème concret, dans l'interaction sociale pour une validation des attributs qui les caractérisent. Dans cette interaction, le tuteur joue un rôle clé complexe de facilitateur, de médiateur, essentiellement par le langage. L'analyse thématique de ses interventions a permis d'en distinguer cinq types: la gestion du groupe, l'argumentation, les questions de différents types, le modelling et les conclusions. Nous avons montré le lien entre les questions du tuteur et le type de réponses des étudiants, pour recommander un meilleur équilibre entre les différents types de questions. Les étudiants, également par les échanges verbaux, mais aussi par la construction collective des cartes conceptuelles initiales et définitives, participent à une co-construction de ces concepts. L'analyse de leurs interactions nous a permis de relever différentes fonctions du langage, pour souligner l'intérêt des interactions argumentatives, marqueurs d'un travail collaboratif en profondeur pour la co-construction des concepts Nous avons aussi montré l'intérêt des cartes conceptuelles non seulement pour visualiser les concepts, mais aussi en tant qu'artefact, outil de médiation psychique à double fonction communicative et sémiotique. Concernant le second objectif, l’exploration du travail personnel des étudiants, on constate que les étudiants de première année font un travail plus approfondi de recherche, et utilisent plus souvent des stratégies de lecture plus efficaces que leurs collègues de deuxième année. Ceux-ci se contentent, en général, des ouvrages de référence, font de simples lectures et s’appuient beaucoup sur les résumés faits par leurs prédécesseurs. Le recours aux ouvrages de référence essentiellement comme source d'information apporte une certaine pauvreté au débat à la phase retour avec peu d'échanges de type argumentatif, témoins d'un travail profond. Ainsi donc, par tout ce soutien qu'elle permet d'apporter aux étudiants pour la construction de leurs connaissances, pour le type d'apprentissage qu'elle offre, l’APP reste une approche unique, digne d’intérêt. Cependant, elle nécessite d'être améliorée par des interventions au niveau du tuteur et des étudiants.

Modelling construction business performance

Searching academic databases for records on ‘business failure’, ‘business distress’ or ‘bankruptcy’ yields a large body of studies on qualitative, empirical, theoretical and simulation aspects. It is a central part of this research to distil from this large quantity of potentially relevant reports and methodologies those which can both flag and predict business failure in the construction industry. An additional search term, such as, ‘construction’, ‘construction industry’ or ‘contractor’ yields a much smaller number of hits, many of which emphasize the construction industry’s distinctive characteristics. We scientists need first to understand the subject of investigation and the environment in which it lives. To do so, an analysis of existing successful and failed approaches to particular research questions is helpful before embarking on new territory. This guides the structure of the following report for we first review papers that specifically report on aspects of business failure in the construction industry followed by, (a) an overview of promising candidates borrowed from other disciplines and industries, and (b) a possible novel approach. An Australian (Queensland) perspective on the topic will also drive this investigation as most of the published research has been applied to the US and UK construction industries.

Analyzing web multimedia query reformulation behavior

Current multimedia Web search engines still use keywords as the primary means to search. Due to the richness in multimedia contents, general users constantly experience some difficulties in formulating textual queries that are representative enough for their needs. As a result, query reformulation becomes part of an inevitable process in most multimedia searches. Previous Web query formulation studies did not investigate the modification sequences and thus can only report limited findings on the reformulation behavior. In this study, we propose an automatic approach to examine multimedia query reformulation using large-scale transaction logs. The key findings show that search term replacement is the most dominant type of modifications in visual searches but less important in audio searches. Image search users prefer the specified search strategy more than video and audio users. There is also a clear tendency to replace terms with synonyms or associated terms in visual queries. The analysis of the search strategies in different types of multimedia searching provides some insights into user’s searching behavior, which can contribute to the design of future query formulation assistance for keyword-based Web multimedia retrieval systems.

Back to basics: Informing the public of co-morbid physical health problems in those with mental illness

Abstract OBJECTIVE: Those with mental illness are at increased risk of physical health problems. The current study aimed to examine the information available online to the Australian public about the increased risk and consequences of physical illness in those with mental health problems and the services available to address these co-morbidities. METHODS: A structured online search was conducted with the search engine Google Australia (www.google.com.au) using generic search terms 'mental health information Australia', 'mental illness information Australia', 'depression', 'anxiety', and 'psychosis'. The direct content of websites was examined for information on the physical co-morbidities of mental illness. All external links on high-profile websites [the first five websites retrieved under each search term (n = 25)] were examined for information pertaining to physical health. RESULTS: Only 4.2% of websites informing the public about mental health contained direct content information about the increased risk of physical co-morbidities. The Australian Government's Department of Health and Ageing site did not contain any information. Of the high-profile websites, 62% had external links to resources about physical health and 55% had recommendations or resources for physical health. Most recommendations were generic. CONCLUSIONS: Relative to the seriousness of this problem, there is a paucity of information available to the public about the increased physical health risks associated with mental illness. Improved public awareness is the starting point of addressing this health inequity.

The influence of OLR1 and PCSK9 gene polymorphisms on ischemic stroke: Evidence from a meta-analysis

It has been reported that both OLR1 and PCSK9 genes are related to various vascular diseases such as atherosclerosis, cardiovascular disease, peripheral artery disease and stroke, in particular ischemic stroke. The prevalence of PCSK9 rs505151 and OLR1 rs11053646 variants in ischemic stroke were 0.005 and 0.116, respectively. However, to date, association between OLR1 rs11053646 and PCSK9 rs505151 polymorphisms and the risk of ischemic stroke remains unclear and inconclusive. Therefore, this first meta-analysis was carried out to clarify the presumed influence of genetic polymorphisms on ischemic stroke, by analyzing the complete coverage of all relevant studies. All eligible case-control and cohort studies that met the search term were retrieved in multiple scientific databases. Data of interest such as demographic data and genotyping methods were extracted from each study, and the meta-analysis was performed using RevMan 5.3 and Metafor R 3.2.1. The pooled odd ratios (ORs) and 95% confidence intervals (CIs) were calculated using both fixed- and random-effect models. A total of seven case-control studies encompassing 1897 ischemic stroke cases and 2119 healthy controls were critically evaluated. Pooled results from the genetic models indicated that OLR1 rs11053646 dominant (OR=1.33. 95%CI:1.11-1.58) and co-dominant models (OR=1.24, 95%CI:1.02-1.51) were significantly associated with ischemic stroke. For PCSK9 rs505151 polymorphism, the OR of co-dominant model (OR=1.36, 95%CI:1.01-1.58) was found to be higher among ischemic stroke patients. In conclusion, the current meta-analysis highlighted that variant allele of OLR1 rs11053646 G>C and PCSK9 rs505151 A>G may contribute to the susceptibility risk of ischemic stroke.

Dimensions of University Student Learning in Medicine and Pharmacy

Study orientations in higher education consist of various dimensions, such as approaches to learning, conceptions of learning and knowledge (i.e. epistemologies), self-regulation, and motivation. They have also been measured in different ways. The main orientations typically reported are reproducing and meaning orientations. The present study explored dimensions of study orientations, focusing in particular on pharmacy and medicine. New versions of self-report instruments were developed and tested in various contexts and in two countries. Furthermore, the linkages between study orientations and students epistemological development were explored. The context of problem-based (PBL) small groups was investigated in order to better understand how collaboration contributes to the quality of learning. The participants of Study I (n=66) were pharmacy students, who were followed during a three-year professionally oriented program in terms of their study orientations and epistemologies. A reproducing orientation to studying diminished during studying, whereas only a few students maintained their original level of meaning orientation. Dualism was found to be associated with a reproducing orientation. In Study II practices associated with deep and surface approaches to learning were measured in two differing ways, in order to better distinguish between what students believed to be useful in studying, and the extent to which they applied their beliefs to practice when preparing for examinations. Differences between domains were investigated by including a sample of Finnish and Swedish medical students (n=956) and a Finnish non-medical sample of university students (n=865). Memorizing and rote learning appeared as differing components of a surface approach to learning, while understanding, relating, and critical evaluation of knowledge emerged as aspects of a deep approach to learning. A structural model confirmed these results in both student samples. Study III explored a wide variety of dimensions of learning in medical education. Swedish medical students (n=280) answered the questionnaire. The deep approach to learning was strongly related to collaboration and reflective learning, whereas the surface approach was associated with novice-like views of knowledge and the valuing of certain and directly applicable knowledge. PBL students aimed at understanding, but also valued the role of memorization. Study IV investigated 12 PBL tutorial groups of students (n=116) studying microbiology and pharmacology in a medical school. The educational application was expected to support a deep approach to learning: Group members course grades in a final examination were related to the perceived functioning of the PBL tutorial groups. Further, the quality of cases that had been used as triggers for learning, was associated with the quality of small group functioning. New dimensions of study orientations were discovered. In particular, novel, finer distinctions were found within the deep approach component. In medicine, critical evaluation of knowledge appeared to be less valued than understanding and relating. Further, collaboration appeared to be closely related to the deep approach, and it was also important in terms of successful PBL studying. The results of the studies confirmed the previously found associations between approaches to learning and study success, but showed interesting context- and subgroup-related differences in this respect. Students ideas about the nature of knowledge and their approaches to learning were shown to be closely related. The present study expanded our understanding of the dimensions of study orientations, of their development, and their contextual variability in pharmacy and medicine.