Thesis (Ph.D.)--University of Wollongong, 2004. / Typescript. Includes bibliographical references: leaf 106-117.

Um dos motivos para o crescente desenvolvimento da área de mineração de dados encontra-se no aumento da quantidade de documentos gerados e armazenados em formato digital, estruturados ou não. A Web contribui sobremaneira para este contexto e, de forma coerente com esta situação, observa-se o surgimento de técnicas específicas para utilização nesta área, como a mineração de estrutura, de conteúdo e de uso. Pode-se afirmar que esta crescente oferta de informação na Web cria o problema da sobrecarga cognitiva. A Hipermídia Adaptativa permite minorar este problema, com a adaptação de hiperdocumentos e hipermídia aos seus usuários segundo suas necessidades, preferências e objetivos. De forma resumida, esta adaptação é realizada relacionando-se informações sobre o domínio da aplicação com informações sobre o perfil de usuários. Um dos tópicos importantes de pesquisa em sistemas de Hipermídia Adaptativa encontra-se na geração e manutenção do perfil dos usuários. Dentre as abordagens conhecidas, existe um contínuo de opções, variando desde cadastros de informações preenchidos manualmente, entrevistas, até a aquisição automática de informações com acompanhamento do uso da Web. Outro ponto fundamental de pesquisa nesta área está ligado à construção das aplicações, sendo que recursos da Web Semântica, como ontologias de domínio ou anotações semânticas de conteúdo podem ser observados no desenvolvimento de sistemas de Hipermídia Adaptativa. Os principais motivos para tal podem ser associados com a inerente flexibilidade, capacidade de compartilhamento e possibilidades de extensão destes recursos. Este trabalho descreve uma arquitetura para a aquisição automática de perfis de classes de usuários, a partir da mineração do uso da Web e da aplicação de ontologias de domínio. O objetivo principal é a integração de informações semânticas, obtidas em uma ontologia de domínio descrevendo o site Web em questão, com as informações de acompanhamento do uso obtidas pela manipulação dos dados de sessões de usuários. Desta forma é possível identificar mais precisamente os interesses e necessidades de um usuário típico. Integra o trabalho a implementação de aplicação de Hipermídia Adaptativa a partir de conceitos de modelagem semântica de aplicações, com a utilização de recursos de serviços Web, para validação experimental da proposta. / One of the reasons for the increasing development observed in Data Mining area is the raising in the quantity of documents generated and stored in digital format, structured or not. The Web plays central role in this context and some specific techniques can be observed, as structure, content and usage mining. This increasing information offer in the Web brings the cognitive overload problem. The Adaptive Hypermedia permits a reduction of this problem, when the contents of selected documents are presented in accordance with the user needs, preferences and objectives. Briefly put, this adaptation is carried out on the basis of relationship between information concerning the application domain and information concerning the user profile. One of the important points in Adaptive Hypermedia systems research is to be found in the generation and maintenance of the user profiles. Some approaches seek to create the user profile from data obtained from registration, others incorporate the results of interviews, and some have the objective of automatic acquisition of information by following the usage. Another fundamental research point is related with the applications construction, where can be observed the use of Web semantic resources, such as semantic annotation and domain ontologies. This work describes the architecture for automatic user profile acquisition, using domain ontologies and Web usage mining. The main objective is the integration of usage data, obtained from user sessions, with semantic description, obtained from a domain ontology. This way it is possible to identify more precisely the interests and needs of a typical user. The implementation of an Adaptive Hypermedia application based on the concepts of semantic application modeling and the use of Web services resources that were integrated into the proposal permitted greater flexibility and experimentation possibilities.

Um dos motivos para o crescente desenvolvimento da área de mineração de dados encontra-se no aumento da quantidade de documentos gerados e armazenados em formato digital, estruturados ou não. A Web contribui sobremaneira para este contexto e, de forma coerente com esta situação, observa-se o surgimento de técnicas específicas para utilização nesta área, como a mineração de estrutura, de conteúdo e de uso. Pode-se afirmar que esta crescente oferta de informação na Web cria o problema da sobrecarga cognitiva. A Hipermídia Adaptativa permite minorar este problema, com a adaptação de hiperdocumentos e hipermídia aos seus usuários segundo suas necessidades, preferências e objetivos. De forma resumida, esta adaptação é realizada relacionando-se informações sobre o domínio da aplicação com informações sobre o perfil de usuários. Um dos tópicos importantes de pesquisa em sistemas de Hipermídia Adaptativa encontra-se na geração e manutenção do perfil dos usuários. Dentre as abordagens conhecidas, existe um contínuo de opções, variando desde cadastros de informações preenchidos manualmente, entrevistas, até a aquisição automática de informações com acompanhamento do uso da Web. Outro ponto fundamental de pesquisa nesta área está ligado à construção das aplicações, sendo que recursos da Web Semântica, como ontologias de domínio ou anotações semânticas de conteúdo podem ser observados no desenvolvimento de sistemas de Hipermídia Adaptativa. Os principais motivos para tal podem ser associados com a inerente flexibilidade, capacidade de compartilhamento e possibilidades de extensão destes recursos. Este trabalho descreve uma arquitetura para a aquisição automática de perfis de classes de usuários, a partir da mineração do uso da Web e da aplicação de ontologias de domínio. O objetivo principal é a integração de informações semânticas, obtidas em uma ontologia de domínio descrevendo o site Web em questão, com as informações de acompanhamento do uso obtidas pela manipulação dos dados de sessões de usuários. Desta forma é possível identificar mais precisamente os interesses e necessidades de um usuário típico. Integra o trabalho a implementação de aplicação de Hipermídia Adaptativa a partir de conceitos de modelagem semântica de aplicações, com a utilização de recursos de serviços Web, para validação experimental da proposta. / One of the reasons for the increasing development observed in Data Mining area is the raising in the quantity of documents generated and stored in digital format, structured or not. The Web plays central role in this context and some specific techniques can be observed, as structure, content and usage mining. This increasing information offer in the Web brings the cognitive overload problem. The Adaptive Hypermedia permits a reduction of this problem, when the contents of selected documents are presented in accordance with the user needs, preferences and objectives. Briefly put, this adaptation is carried out on the basis of relationship between information concerning the application domain and information concerning the user profile. One of the important points in Adaptive Hypermedia systems research is to be found in the generation and maintenance of the user profiles. Some approaches seek to create the user profile from data obtained from registration, others incorporate the results of interviews, and some have the objective of automatic acquisition of information by following the usage. Another fundamental research point is related with the applications construction, where can be observed the use of Web semantic resources, such as semantic annotation and domain ontologies. This work describes the architecture for automatic user profile acquisition, using domain ontologies and Web usage mining. The main objective is the integration of usage data, obtained from user sessions, with semantic description, obtained from a domain ontology. This way it is possible to identify more precisely the interests and needs of a typical user. The implementation of an Adaptive Hypermedia application based on the concepts of semantic application modeling and the use of Web services resources that were integrated into the proposal permitted greater flexibility and experimentation possibilities.

No description available.

No description available.

Web Usage Mining (WUM) can be used to determine whether the information architecture of a web site is structured correctly. Existing WUM tools however, do not indicate which web usage mining algorithms are used or provide effective graphical visualisations of the results obtained. WUM techniques can be used to determine typical navigation patterns of the users of organisational web sites. An organisational web site can be described as a site which has a high level of content. The Computer Science & Information Systems (CS&IS) web site at the Nelson Mandela Metropolitan University (NMMU) is an example of such a web site. The process of combining WUM and information visualisation techniques in order to discover useful information about web usage patterns is called visual web mining. The goal of this research is to discuss the development of a WUM model and a prototype, called WebPatterns, which allows the user to effectively visualise web usage patterns of an organisational web site. This will facilitate determining whether the information architecture of the CS&IS web site is structured correctly. The WUM algorithms used in WebPatterns are association rule mining and sequence analysis. The purpose of association rule mining is to discover relationships between different web pages within a web site. Sequence analysis is used to determine the longest time ordered paths that satisfy a user specified minimum frequency. A radial tree layout is used in WebPatterns to visualise the static structure of the organisational web site. The structure of the web site is laid out radially, with the home page in the middle and other pages positioned in circles at various levels around it. Colour and other visual cues are used to show the results of the WUM algorithms. User testing was used to determine the effectiveness and usefulness of WebPatterns for visualising web usage patterns. The results of the user testing clearly show that the participants were highly satisfied with the visual design and information provided by WebPatterns. All the participants also indicated that they would like to use WebPatterns in the future. Analysis of the web usage patterns presented by WebPatterns was used to determine that the information architecture of the CS&IS web site can be restructured to better facilitate information retrieval. Changes to the CS&IS web site web were suggested, included placing embedded hyperlinks on the home page to the frequently accessed sections of the web site.

General characteristic of web mining including methodology and procedures incorporated into this term. Relation to other areas (data mining, artificial intelligence, statistics, databases, internet technologies, management etc.) Web usage mining - data sources, data pre-processing, characterization of analytical methods and tools, interpretation of outputs (results), and possible areas of usage including examples. Suggestion of solution method, realization and a concrete example's outputs interpretation while using above mentioned methods of web usage mining.

This master thesis aims to investigate the possibilities of predicting purchase intentions of customers during their sales processes in the real estate sector. Also, the web activity of customers on a real estate company’s web site is used as the basis for the forecasting. A machine learning framework has been developed, where its compliance with the GDPR is also assessed.  Five supervised machine learning algorithms – logistic regression, k-nearest neighbors, decision tree, random forest, multilayer perceptron – have been utilized for predicting the classes of the customers: buyers and non-buyers. Three data sets were generated, which represented the total number of active customers at different points in time: at the same day as a sales process starts (day 0) and 10 and 20 days after it. The algorithms were applied and evaluated on these data sets to identify when it is suitable to predict the purchase intentions of customers. To increase the generalization capability of the algorithms, hyperparameter optimization along with data resampling by combining undersampling and synthetic minority over-sampling techniques, k-fold cross validation and mutual information, as feature selection, were applied. The results show that the number of visited web pages, sessions, searched projects (concerning accommodations) and searched locations were relevant for all three data sets. The average price (in total and per square meter) of the most frequently visited web page regarding projects were also included in all the data sets. In addition, the total number of registration of interests sent, and the total amount of time spent on the company’s web site were considered in the second (day 10) and third data set (day 20). Further, a multilayer perceptron – applied 10 days after the start of a sales process – was considered as the optimal model for classifying the purchase intentions of customers. Moreover, the developed machine learning framework is argued to be compliant with the GDPR. Further evaluation regarding the compliance needs to be conducted if the methodology of this machine learning framework would be implemented in practice.

No description available.

Given set of web pages labeled with ontological items, the level of similarity between two web pages is measured using the level of similarity between ontological items of pages labeled with. Using similarity measure between two pages, degree of similarity between two sequences of web page visits can be calculated as well. Using clustering algorithms, similar frequent sequences are grouped and representative sequences are selected from these groups. A new sequence is compared with all clusters and it is assigned to most similar one. Representatives of the most similar cluster can be used in several real world cases. They can be used for predicting and prefetching the next page user will visit or for helping the navigation of user in the website. They can also be used to improve the structure of website for easier navigation. In this study the effect of time spent on each web page during the session is analyzed.