Learning representations for Information Retrieval

La recherche d'informations s'intéresse, entre autres, à répondre à des questions comme: est-ce qu'un document est pertinent à une requête ? Est-ce que deux requêtes ou deux documents sont similaires ? Comment la similarité entre deux requêtes ou documents peut être utilisée pour améliorer l'estimation de la pertinence ? Pour donner réponse à ces questions, il est nécessaire d'associer chaque document et requête à des représentations interprétables par ordinateur. Une fois ces représentations estimées, la similarité peut correspondre, par exemple, à une distance ou une divergence qui opère dans l'espace de représentation. On admet généralement que la qualité d'une représentation a un impact direct sur l'erreur d'estimation par rapport à la vraie pertinence, jugée par un humain. Estimer de bonnes représentations des documents et des requêtes a longtemps été un problème central de la recherche d'informations. Le but de cette thèse est de proposer des nouvelles méthodes pour estimer les représentations des documents et des requêtes, la relation de pertinence entre eux et ainsi modestement avancer l'état de l'art du domaine. Nous présentons quatre articles publiés dans des conférences internationales et un article publié dans un forum d'évaluation. Les deux premiers articles concernent des méthodes qui créent l'espace de représentation selon une connaissance à priori sur les caractéristiques qui sont importantes pour la tâche à accomplir. Ceux-ci nous amènent à présenter un nouveau modèle de recherche d'informations qui diffère des modèles existants sur le plan théorique et de l'efficacité expérimentale. Les deux derniers articles marquent un changement fondamental dans l'approche de construction des représentations. Ils bénéficient notamment de l'intérêt de recherche dont les techniques d'apprentissage profond par réseaux de neurones, ou deep learning, ont fait récemment l'objet. Ces modèles d'apprentissage élicitent automatiquement les caractéristiques importantes pour la tâche demandée à partir d'une quantité importante de données. Nous nous intéressons à la modélisation des relations sémantiques entre documents et requêtes ainsi qu'entre deux ou plusieurs requêtes. Ces derniers articles marquent les premières applications de l'apprentissage de représentations par réseaux de neurones à la recherche d'informations. Les modèles proposés ont aussi produit une performance améliorée sur des collections de test standard. Nos travaux nous mènent à la conclusion générale suivante: la performance en recherche d'informations pourrait drastiquement être améliorée en se basant sur les approches d'apprentissage de représentations.

Learning representations for Information Retrieval

La recherche d'informations s'intéresse, entre autres, à répondre à des questions comme: est-ce qu'un document est pertinent à une requête ? Est-ce que deux requêtes ou deux documents sont similaires ? Comment la similarité entre deux requêtes ou documents peut être utilisée pour améliorer l'estimation de la pertinence ? Pour donner réponse à ces questions, il est nécessaire d'associer chaque document et requête à des représentations interprétables par ordinateur. Une fois ces représentations estimées, la similarité peut correspondre, par exemple, à une distance ou une divergence qui opère dans l'espace de représentation. On admet généralement que la qualité d'une représentation a un impact direct sur l'erreur d'estimation par rapport à la vraie pertinence, jugée par un humain. Estimer de bonnes représentations des documents et des requêtes a longtemps été un problème central de la recherche d'informations. Le but de cette thèse est de proposer des nouvelles méthodes pour estimer les représentations des documents et des requêtes, la relation de pertinence entre eux et ainsi modestement avancer l'état de l'art du domaine. Nous présentons quatre articles publiés dans des conférences internationales et un article publié dans un forum d'évaluation. Les deux premiers articles concernent des méthodes qui créent l'espace de représentation selon une connaissance à priori sur les caractéristiques qui sont importantes pour la tâche à accomplir. Ceux-ci nous amènent à présenter un nouveau modèle de recherche d'informations qui diffère des modèles existants sur le plan théorique et de l'efficacité expérimentale. Les deux derniers articles marquent un changement fondamental dans l'approche de construction des représentations. Ils bénéficient notamment de l'intérêt de recherche dont les techniques d'apprentissage profond par réseaux de neurones, ou deep learning, ont fait récemment l'objet. Ces modèles d'apprentissage élicitent automatiquement les caractéristiques importantes pour la tâche demandée à partir d'une quantité importante de données. Nous nous intéressons à la modélisation des relations sémantiques entre documents et requêtes ainsi qu'entre deux ou plusieurs requêtes. Ces derniers articles marquent les premières applications de l'apprentissage de représentations par réseaux de neurones à la recherche d'informations. Les modèles proposés ont aussi produit une performance améliorée sur des collections de test standard. Nos travaux nous mènent à la conclusion générale suivante: la performance en recherche d'informations pourrait drastiquement être améliorée en se basant sur les approches d'apprentissage de représentations.

Learning mathematics in a classroom community of inquiry

This article considers the question of what specific actions a teacher might take to create a culture of inquiry in a secondary school mathematics classroom. Sociocultural theories of learning provide the framework for examining teaching and learning practices in a single classroom over a two-year period. The notion of the zone of proximal development (ZPD) is invoked as a fundamental framework for explaining learning as increasing participation in a community of practice characterized by mathematical inquiry. The analysis draws on classroom observation and interviews with students and the teacher to show how the teacher established norms and practices that emphasized mathematical sense-making and justification of ideas and arguments and to illustrate the learning practices that students developed in response to these expectations.

Learning task specific distributed paragraph representations using a 2-tier convolutional neural network

We introduce a type of 2-tier convolutional neural network model for learning distributed paragraph representations for a special task (e.g. paragraph or short document level sentiment analysis and text topic categorization). We decompose the paragraph semantics into 3 cascaded constitutes: word representation, sentence composition and document composition. Specifically, we learn distributed word representations by a continuous bag-of-words model from a large unstructured text corpus. Then, using these word representations as pre-trained vectors, distributed task specific sentence representations are learned from a sentence level corpus with task-specific labels by the first tier of our model. Using these sentence representations as distributed paragraph representation vectors, distributed paragraph representations are learned from a paragraph-level corpus by the second tier of our model. It is evaluated on DBpedia ontology classification dataset and Amazon review dataset. Empirical results show the effectiveness of our proposed learning model for generating distributed paragraph representations.

Interface Engineering and Design: Adaptibility Problems

* The research work reviewed in this paper has been carried out in the context of the Russian Foundation for Basic Research funded project “Adaptable Intelligent Interfaces Research and Development for Distance Learning Systems”(grant N 02-01-81019). The authors wish to acknowledge the co-operation with the Byelorussian partners of this project.

A gloss composition and context clustering based distributed word sense representation model

In recent years, there has been an increasing interest in learning a distributed representation of word sense. Traditional context clustering based models usually require careful tuning of model parameters, and typically perform worse on infrequent word senses. This paper presents a novel approach which addresses these limitations by first initializing the word sense embeddings through learning sentence-level embeddings from WordNet glosses using a convolutional neural networks. The initialized word sense embeddings are used by a context clustering based model to generate the distributed representations of word senses. Our learned representations outperform the publicly available embeddings on half of the metrics in the word similarity task, 6 out of 13 sub tasks in the analogical reasoning task, and gives the best overall accuracy in the word sense effect classification task, which shows the effectiveness of our proposed distributed distribution learning model.

Game-based Approach in e-Learning

Report published in the Proceedings of the National Conference on "Education and Research in the Information Society", Plovdiv, May, 2016

Development of an integrated surface and subsurface model of Everglades National Park

An integrated surface-subsurface hydrological model of Everglades National Park (ENP) was developed using MIKE SHE and MIKE 11 modeling software. The model has a resolution of 400 meters, covers approximately 1050 square miles of ENP, includes 110 miles of drainage canals with a variety of hydraulic structures, and processes hydrological information, such as evapotranspiration, precipitation, groundwater levels, canal discharges and levels, and operational schedules. Calibration was based on time series and probability of exceedance for water levels and discharges in the years 1987 through 1997. Model verification was then completed for the period of 1998 through 2005. Parameter sensitivity in uncertainty analysis showed that the model was most sensitive to the hydraulic conductivity of the regional Surficial Aquifer System, the Manning's roughness coefficient, and the leakage coefficient, which defines the canal-subsurface interaction. The model offers an enhanced predictive capability, compared to other models currently available, to simulate the flow regime in ENP and to forecast the impact of topography, water flows, and modifying operation schedules.

Increasing Informal Learning Opportunities for Teachers in the School Setting

This paper comprehensively defines how to implement informal learning strategies into the classroom setting using Marsick and Watkins’s Incidental Learning Model (2001). Existing barriers that stand between educators and informal learning in the school setting are explained. Implications for removing said inhibitors while increasing learning are explicated.

Development of a work-based learning MSc course which incorporates the development and demonstration of professional engineering competence standards

UK engineering standards are regulated by the Engineering Council (EC) using a set of generic threshold competence standards which all professionally registered Chartered Engineers in the UK must demonstrate, underpinned by a separate academic qualification at Masters Level. As part of an EC-led national project for the development of work-based learning (WBL) courses leading to Chartered Engineer registration, Aston University has started an MSc Professional Engineering programme, a development of a model originally designed by Kingston University, and build around a set of generic modules which map onto the competence standards. The learning pedagogy of these modules conforms to a widely recognised experiential learning model, with refinements incorporated from a number of other learning models. In particular, the use of workplace mentoring to support the development of critical reflection and to overcome barriers to learning is being incorporated into the learning space. This discussion paper explains the work that was done in collaboration with the EC and a number of Professional Engineering Institutions, to design a course structure and curricular framework that optimises the engineering learning process for engineers already working across a wide range of industries, and to address issues of engineering sustainability. It also explains the thinking behind the work that has been started to provide an international version of the course, built around a set of globalised engineering competences. © 2010 W J Glew, E F Elsworth.

What Makes Online Grocers Work? A Case Study Analysis of Factors Contributing to Online Grocery Store Profitability

This research examined the factors contributing to the performance of online grocers prior to, and following, the 2000 dot.com collapse. The primary goals were to assess the relationship between a company’s business model(s) and its performance in the online grocery channel and to determine if there were other company and/or market related factors that could account for company performance. To assess the primary goals, a case based theory building process was utilized. A three-way cross-case analysis comprising Peapod, GroceryWorks, and Tesco examined the common profit components, the structural category (e.g., pure-play, partnership, and hybrid) profit components, and the idiosyncratic profit components related to each specific company. Based on the analysis, it was determined that online grocery store business models could be represented at three distinct, but hierarchically, related levels. The first level was termed the core model and represented the basic profit structure that all online grocers needed in order to conduct operations. The next model level was termed the structural model and represented the profit structure associated with the specific business model configuration (i.e., pure-play, partnership, hybrid). The last model level was termed the augmented model and represented the company’s business model when idiosyncratic profit components were included. In relation to the five company related factors, scalability, rate of expansion, and the automation level were potential candidates for helping to explain online grocer performance. In addition, all the market structure related factors were deemed possible candidates for helping to explain online grocer performance. The study concluded by positing an alternative hypothesis concerning the performance of online grocers. Prior to this study, the prevailing wisdom was that the business models were the primary cause of online grocer performance. However, based on the core model analysis, it was hypothesized that the customer relationship activities (i.e., advertising, promotions, and loyalty program tie-ins) were the real drivers of online grocer performance.

Extracting Patterns from Educational Traces via Clustering and Associated Quality Metrics

Clustering algorithms, pattern mining techniques and associated quality metrics emerged as reliable methods for modeling learners’ performance, comprehension and interaction in given educational scenarios. The specificity of available data such as missing values, extreme values or outliers, creates a challenge to extract significant user models from an educational perspective. In this paper we introduce a pattern detection mechanism with-in our data analytics tool based on k-means clustering and on SSE, silhouette, Dunn index and Xi-Beni index quality metrics. Experiments performed on a dataset obtained from our online e-learning platform show that the extracted interaction patterns were representative in classifying learners. Furthermore, the performed monitoring activities created a strong basis for generating automatic feedback to learners in terms of their course participation, while relying on their previous performance. In addition, our analysis introduces automatic triggers that highlight learners who will potentially fail the course, enabling tutors to take timely actions.

Student attitudes towards cooperative learning in education

Abstract: Active or participatory learning by the student within a classroom environment has been fairly recently recognized as an effective, efficient, and superior instructional technique yet few teachers in higher education have adopted this pedagogical strategy. This is especially true in Science where teachers primarily lecture to passively seated students while using static visual aids or multimedia projections. Teachers generally teach as they were taught and lecture formats have been the norm. Although student-learning theories as well as student learning styles, abilities, and understanding strategies have changed, traditional teaching techniques have not evolved past the “chalk and talk” instructional strategy. This research looked into student’s perceptions of cooperative learning or team-based active learning in order to gain insight and some understanding as to how students felt about this learning technique. Student’s attitudes were then compared to student grades to detennine whether cooperative learning impeded or ameliorated academic performance. The results revealed significant differences measured in all the survey questions pertaining to perception or attitudes. As a result of the cooperative learning activities, respondents indicated more agreement to the survey questions pertaining to the benefits of cooperative learning. The experimental group exposed to cooperative learning thus experienced more positive attitudes and perceptions than the groups exposed only to a lecture-based teaching and learning format. Each of the hypotheses tested demonstrated that students had more positive attitudes towards cooperative learning strategies. Recommendations as to future work were presented in order to gain a greater understanding into both student and teacher attitudes towards the cooperative learning model.||Résumé: Lapprentissage actif ou préparatoire par létudiant au sein d’une classe a été reconnu assez récemment comme une technique d’enseignement plus efficace. Cependant, peu d’enseignants ont adopté cette stratégie pedagogique pour l'éducation post-secondaire. Ceci est particulièrement le cas dans le domaine des sciences où les enseignants font surtout usage de cours magistraux avec des étudiants passifs tout en utilisant des aides visuelles statiques ou des projections multimédias. Les professeurs enseignent generalement comme on leur a eux-même enseigné et les cours magistraux ont été la norme par le passé. Les techniques traditionnelles d'enseignernent n'ont pas évolué au-delà de la craie et du tableau noir et ce même si les théories sur l’apprentissage par les étudiants ont changé, tout comme les styles, les habiletés et les stratégies de compréhension d’apprentissage des étudiants. Cette recherche se penche sur les perceptions des étudiants au sujet de l'apprentissage coopératif ou de l'apprentissage actif par équipe de telle sorte qu'on puisse avoir un aperçu et une certaine compréhension de comment les étudiants se sentent par rapport à ces techniques d'apprentissage. Les attitudes des étudiants ont par la suite été comparées aux notes de ceux-ci pour déterminer si l'apprentissage coopératif avait nui ou au contraire amélioré leurs performances académiques. Les résultats obtenus dans l'étude d'ensemble révèlent des différences significatives dans toutes les questions ayant trait à la perception et aux attitudes.

English syllable confusion and imitation in Korean bilingual and monolingual children and adults

In the present study, Korean-English bilingual (KEB) and Korean monolingual (KM) children, between the ages of 8 and 13 years, and KEB adults, ages 18 and older, were examined with one speech perception task, called the Nonsense Syllable Confusion Matrix (NSCM) task (Allen, 2005), and two production tasks, called the Nonsense Syllable Imitation Task (NSIT) and the Nonword Repetition Task (NRT; Dollaghan & Campbell, 1998). The present study examined (a) which English sounds on the NSCM task were identified less well, presumably due to interference from Korean phonology, in bilinguals learning English as a second language (L2) and in monolinguals learning English as a foreign language (FL); (b) which English phonemes on the NSIT were more challenging for bilinguals and monolinguals to produce; (c) whether perception on the NSCM task is related to production on the NSIT, or phonological awareness, as measured by the NRT; and (d) whether perception and production differ in three age-language status groups (i.e., KEB children, KEB adults, and KM children) and in three proficiency subgroups of KEB children (i.e., English-dominant, ED; balanced, BAL; and Korean-dominant, KD). In order to determine English proficiency in each group, language samples were extensively and rigorously analyzed, using software, called Systematic Analysis of Language Transcripts (SALT). Length of samples in complete and intelligible utterances, number of different and total words (NDW and NTW, respectively), speech rate in words per minute (WPM), and number of grammatical errors, mazes, and abandoned utterances were measured and compared among the three initial groups and the three proficiency subgroups. Results of the language sample analysis (LSA) showed significant group differences only between the KEBs and the KM children, but not between the KEB children and adults. Nonetheless, compared to normative means (from a sample length- and age-matched database provided by SALT), the KEB adult group and the KD subgroup produced English at significantly slower speech rates than expected for monolingual, English-speaking counterparts. Two existing models of bilingual speech perception and production—the Speech Learning Model or SLM (Flege, 1987, 1992) and the Perceptual Assimilation Model or PAM (Best, McRoberts, & Sithole, 1988; Best, McRoberts, & Goodell, 2001)—were considered to see if they could account for the perceptual and production patterns evident in the present study. The selected English sounds for stimuli in the NSCM task and the NSIT were 10 consonants, /p, b, k, g, f, θ, s, z, ʧ, ʤ/, and 3 vowels /I, ɛ, æ/, which were used to create 30 nonsense syllables in a consonant-vowel structure. Based on phonetic or phonemic differences between the two languages, English sounds were categorized either as familiar sounds—namely, English sounds that are similar, but not identical, to L1 Korean, including /p, k, s, ʧ, ɛ/—or unfamiliar sounds—namely, English sounds that are new to L1, including /b, g, f, θ, z, ʤ, I, æ/. The results of the NSCM task showed that (a) consonants were perceived correctly more often than vowels, (b) familiar sounds were perceived correctly more often than unfamiliar ones, and (c) familiar consonants were perceived correctly more often than unfamiliar ones across the three age-language status groups and across the three proficiency subgroups; and (d) the KEB children perceived correctly more often than the KEB adults, the KEB children and adults perceived correctly more often than the KM children, and the ED and BAL subgroups perceived correctly more often than the KD subgroup. The results of the NSIT showed (a) consonants were produced more accurately than vowels, and (b) familiar sounds were produced more accurately than unfamiliar ones, across the three age-language status groups. Also, (c) familiar consonants were produced more accurately than unfamiliar ones in the KEB and KM child groups, and (d) unfamiliar vowels were produced more accurately than a familiar one in the KEB child group, but the reverse was true in the KEB adult and KM child groups. The KEB children produced sounds correctly significantly more often than the KM children and the KEB adults, though the percent correct differences were smaller than for perception. Production differences were not found among the three proficiency subgroups. Perception on the NSCM task was compared to production on the NSIT and NRT. Weak positive correlations were found between perception and production (NSIT) for unfamiliar consonants and sounds, whereas a weak negative correlation was found for unfamiliar vowels. Several correlations were significant for perceptual performance on the NSCM task and overall production performance on the NRT: for unfamiliar consonants, unfamiliar vowels, unfamiliar sounds, consonants, vowels, and overall performance on the NSCM task. Nonetheless, no significant correlation was found between production on the NSIT and NRT. Evidently these are two very different production tasks, where immediate imitation of single syllables on the NSIT results in high performance for all groups. Findings of the present study suggest that (a) perception and production of L2 consonants differ from those of vowels; (b) perception and production of L2 sounds involve an interaction of sound type and familiarity; (c) a weak relation exists between perception and production performance for unfamiliar sounds; and (d) L2 experience generally predicts perceptual and production performance. The present study yields several conclusions. The first is that familiarity of sounds is an important influence on L2 learning, as claimed by both SLM and PAM. In the present study, familiar sounds were perceived and produced correctly more often than unfamiliar ones in most cases, in keeping with PAM, though experienced L2 learners (i.e., the KEB children) produced unfamiliar vowels better than familiar ones, in keeping with SLM. Nonetheless, the second conclusion is that neither SLM nor PAM consistently and thoroughly explains the results of the present study. This is because both theories assume that the influence of L1 on the perception of L2 consonants and vowels works in the same way as for production of them. The third and fourth conclusions are two proposed arguments: that perception and production of consonants are different than for vowels, and that sound type interacts with familiarity and L2 experience. These two arguments can best explain the current findings. These findings may help us to develop educational curricula for bilingual individuals listening to and articulating English. Further, the extensive analysis of spontaneous speech in the present study should contribute to the specification of parameters for normal language development and function in Korean-English bilingual children and adults.