Augmenting web service discovery by cognitive semantics and abduction

We argue that web service discovery technology should help the user navigate a complex problem space by providing suggestions for services which they may not be able to formulate themselves as (s)he lacks the epistemic resources to do so. Free text documents in service environments provide an untapped source of information for augmenting the epistemic state of the user and hence their ability to search effectively for services. A quantitative approach to semantic knowledge representation is adopted in the form of semantic space models computed from these free text documents. Knowledge of the user’s agenda is promoted by associational inferences computed from the semantic space. The inferences are suggestive and aim to promote human abductive reasoning to guide the user from fuzzy search goals into a better understanding of the problem space surrounding the given agenda. Experimental results are discussed based on a complex and realistic planning activity.

Similarity metrics within a point of view

In vector space based approaches to natural language processing, similarity is commonly measured by taking the angle between two vectors representing words or documents in a semantic space. This is natural from a mathematical point of view, as the angle between unit vectors is, up to constant scaling, the only unitarily invariant metric on the unit sphere. However, similarity judgement tasks reveal that human subjects fail to produce data which satisfies the symmetry and triangle inequality requirements for a metric space. A possible conclusion, reached in particular by Tversky et al., is that some of the most basic assumptions of geometric models are unwarranted in the case of psychological similarity, a result which would impose strong limits on the validity and applicability vector space based (and hence also quantum inspired) approaches to the modelling of cognitive processes. This paper proposes a resolution to this fundamental criticism of of the applicability of vector space models of cognition. We argue that pairs of words imply a context which in turn induces a point of view, allowing a subject to estimate semantic similarity. Context is here introduced as a point of view vector (POVV) and the expected similarity is derived as a measure over the POVV's. Different pairs of words will invoke different contexts and different POVV's. Hence the triangle inequality ceases to be a valid constraint on the angles. We test the proposal on a few triples of words and outline further research.

Modelling word meaning using efficient tensor representations

Models of word meaning, built from a corpus of text, have demonstrated success in emulating human performance on a number of cognitive tasks. Many of these models use geometric representations of words to store semantic associations between words. Often word order information is not captured in these models. The lack of structural information used by these models has been raised as a weakness when performing cognitive tasks. This paper presents an efficient tensor based approach to modelling word meaning that builds on recent attempts to encode word order information, while providing flexible methods for extracting task specific semantic information.

Projecting Computational Sense of Self: A Study of Transition in a Chronic Illness Online Community

We report on analysis of discussions in an online community of people with chronic illness using socio-cognitively motivated, automatically produced semantic spaces. The analysis aims to further the emerging theory of "transition" (how people can learn to incorporate the consequences of illness into their lives). An automatically derived representation of sense of self for individuals is created in the semantic space by the analysis of the email utterances of the community members. The movement over time of the sense of self is visualised, via projection, with respect to axes of "ordinariness" and "extra-ordinariness". Qualitative evaluation shows that the visualisation is paralleled by the transitions of people during the course of their illness. The research aims to progress tools for analysis of textual data to promote greater use of tacit knowledge as found in online virtual communities. We hope it also encourages further interest in representation of sense-of-self.

A tensor encoding model of word meaning : theory and application to information retrieval

This project was a step forward in developing and evaluating a novel, mathematical model that can deduce the meaning of words based on their use in language. This model can be applied to a wide range of natural language applications, including the information seeking process most of us undertake on a daily basis.

Some extensions to representation and encoding of structure in models of distributional semantics

This thesis makes several contributions towards improved methods for encoding structure in computational models of word meaning. New methods are proposed and evaluated which address the requirement of being able to easily encode linguistic structural features within a computational representation while retaining the ability to scale to large volumes of textual data. Various methods are implemented and evaluated on a range of evaluation tasks to demonstrate the effectiveness of the proposed methods.

The heterogeneous cluster ensemble method using hubness for clustering text documents

We propose a cluster ensemble method to map the corpus documents into the semantic space embedded in Wikipedia and group them using multiple types of feature space. A heterogeneous cluster ensemble is constructed with multiple types of relations i.e. document-term, document-concept and document-category. A final clustering solution is obtained by exploiting associations between document pairs and hubness of the documents. Empirical analysis with various real data sets reveals that the proposed meth-od outperforms state-of-the-art text clustering approaches.

Comprehending adverbs of doubt and certainty in health communication: A multidimensional scaling approach

This research explored the feasibility of using multidimensional scaling (MDS) analysis in novel combination with other techniques to study comprehension of epistemic adverbs expressing doubt and certainty (e.g., evidently, obviously, probably) as they relate to health communication in clinical settings. In Study 1, Australian English speakers performed a dissimilarity-rating task with sentence pairs containing the target stimuli, presented as "doctors' opinions". Ratings were analyzed using a combination of cultural consensus analysis (factor analysis across participants), weighted-data classical-MDS, and cluster analysis. Analyses revealed strong within-community consistency for a 3-dimensional semantic space solution that took into account individual differences, strong statistical acceptability of the MDS results in terms of stress and explained variance, and semantic configurations that were interpretable in terms of linguistic analyses of the target adverbs. The results confirmed the feasibility of using MDS in this context. Study 2 replicated the results with Canadian English speakers on the same task. Semantic analyses and stress decomposition analysis were performed on the Australian and Canadian data sets, revealing similarities and differences between the two groups. Overall, the results support using MDS to study comprehension of words critical for health communication, including in future studies, for example, second language speaking patients and/or practitioners. More broadly, the results indicate that the techniques described should be promising for comprehension studies in many communicative domains, in both clinical settings and beyond, and including those targeting other aspects of language and focusing on comparisons across different speech communities.

Déploiement automatique d’une application de routage téléphonique d’une langue source vers une langue cible

Les modèles de compréhension statistiques appliqués à des applications vocales nécessitent beaucoup de données pour être entraînés. Souvent, une même application doit pouvoir supporter plusieurs langues, c’est le cas avec les pays ayant plusieurs langues officielles. Il s’agit donc de gérer les mêmes requêtes des utilisateurs, lesquelles présentent une sémantique similaire, mais dans plusieurs langues différentes. Ce projet présente des techniques pour déployer automatiquement un modèle de compréhension statistique d’une langue source vers une langue cible. Ceci afin de réduire le nombre de données nécessaires ainsi que le temps relié au déploiement d’une application dans une nouvelle langue. Premièrement, une approche basée sur les techniques de traduction automatique est présentée. Ensuite une approche utilisant un espace sémantique commun pour comparer plusieurs langues a été développée. Ces deux méthodes sont comparées pour vérifier leurs limites et leurs faisabilités. L’apport de ce projet se situe dans l’amélioration d’un modèle de traduction grâce à l’ajout de données très proche de l’application ainsi que d’une nouvelle façon d’inférer un espace sémantique multilingue.

Conceptualisation et communication des nuances de timbre à la guitare classique

La version intégrale de cette thèse est disponible uniquement pour consultation individuelle à la Bibliothèque de musique de l’Université de Montréal (www.bib.umontreal.ca/MU).

Apprentissage d'espaces sémantiques

Dans cette dissertation, nous présentons plusieurs techniques d’apprentissage d’espaces sémantiques pour plusieurs domaines, par exemple des mots et des images, mais aussi à l’intersection de différents domaines. Un espace de représentation est appelé sémantique si des entités jugées similaires par un être humain, ont leur similarité préservée dans cet espace. La première publication présente un enchaînement de méthodes d’apprentissage incluant plusieurs techniques d’apprentissage non supervisé qui nous a permis de remporter la compétition “Unsupervised and Transfer Learning Challenge” en 2011. Le deuxième article présente une manière d’extraire de l’information à partir d’un contexte structuré (177 détecteurs d’objets à différentes positions et échelles). On montrera que l’utilisation de la structure des données combinée à un apprentissage non supervisé permet de réduire la dimensionnalité de 97% tout en améliorant les performances de reconnaissance de scènes de +5% à +11% selon l’ensemble de données. Dans le troisième travail, on s’intéresse à la structure apprise par les réseaux de neurones profonds utilisés dans les deux précédentes publications. Plusieurs hypothèses sont présentées et testées expérimentalement montrant que l’espace appris a de meilleures propriétés de mixage (facilitant l’exploration de différentes classes durant le processus d’échantillonnage). Pour la quatrième publication, on s’intéresse à résoudre un problème d’analyse syntaxique et sémantique avec des réseaux de neurones récurrents appris sur des fenêtres de contexte de mots. Dans notre cinquième travail, nous proposons une façon d’effectuer de la recherche d’image ”augmentée” en apprenant un espace sémantique joint où une recherche d’image contenant un objet retournerait aussi des images des parties de l’objet, par exemple une recherche retournant des images de ”voiture” retournerait aussi des images de ”pare-brises”, ”coffres”, ”roues” en plus des images initiales.

Never the twain shall meet: A memetic analysis of poverty perceptions

Poverty, as defined within development discourse, does not fully capture the reality in which the poor live, which is formed also by values and beliefs specific to a given culture and setting. This article uses a memetic approach to investigating the reality of poverty among pastoralists and urban dwellers in Kenya. By distinguishing the semantic space and the cultural context in which the definitions are framed, it enables the researcher to make sufficient generalisations while also recognising the differences between cultures. The results demonstrate how pastoralists and urban dwellers conceptualise poverty differently particularly in regard to causes. Further, the article suggests that development actors often utilise a Western construct which does not entirely reflect the values and beliefs of the poor.

The shape of change: memetic analysis of poverty definitions from the 1970s to the 2000s

The authors illustrate how notions of poverty are constructed around specific ‘memes’, or replicating units of cultural information, around which concepts and ideas develop and change. Three ‘memes’ characterising definitions of poverty over the previous years were identified: ‘basic needs’, ‘multidimensional’ and ‘deprivation’. The analysis illustrated the semantic space in which each term was utilised and to the extent it changed and modified over time by different actors. The results revealed how ‘memes’ compete with one another across the discourse. Within this competition, older concepts are almost never fully abandoned, but rather repackaged and reutilised. Thus, new definitions of poverty are less innovative than portrayed in the wider literature.

Design e semântica: investigação de técnicas estatísticas para auxílio no projeto de produto

Coordenação de Aperfeiçoamento de Pessoal de Nível Superior (CAPES)

Information Processing in a Cognitive Model of NLP

A model of the cognitive process of natural language processing has been developed using the formalism of generalized nets. Following this stage-simulating model, the treatment of information inevitably includes phases, which require joint operations in two knowledge spaces – language and semantics. In order to examine and formalize the relations between the language and the semantic levels of treatment, the language is presented as an information system, conceived on the bases of human cognitive resources, semantic primitives, semantic operators and language rules and data. This approach is applied for modeling a specific grammatical rule – the secondary predication in Russian. Grammatical rules of the language space are expressed as operators in the semantic space. Examples from the linguistics domain are treated and several conclusions for the semantics of the modeled rule are made. The results of applying the information system approach to the language turn up to be consistent with the stages of treatment modeled with the generalized net.