Hybrid Approach Combining Machine Learning and a Rule-Based Expert System for Text Categorization

This paper discusses a novel hybrid approach for text categorization that combines a machine learning algorithm, which provides a base model trained with a labeled corpus, with a rule-based expert system, which is used to improve the results provided by the previous classifier, by filtering false positives and dealing with false negatives. The main advantage is that the system can be easily fine-tuned by adding specific rules for those noisy or conflicting categories that have not been successfully trained. We also describe an implementation based on k-Nearest Neighbor and a simple rule language to express lists of positive, negative and relevant (multiword) terms appearing in the input text. The system is evaluated in several scenarios, including the popular Reuters-21578 news corpus for comparison to other approaches, and categorization using IPTC metadata, EUROVOC thesaurus and others. Results show that this approach achieves a precision that is comparable to top ranked methods, with the added value that it does not require a demanding human expert workload to train

Text categorization methods for automatic estimation of verbal intelligence

In this paper we investigate whether conventional text categorization methods may suffice to infer different verbal intelligence levels. This research goal relies on the hypothesis that the vocabulary that speakers make use of reflects their verbal intelligence levels. Automatic verbal intelligence estimation of users in a spoken language dialog system may be useful when defining an optimal dialog strategy by improving its adaptation capabilities. The work is based on a corpus containing descriptions (i.e. monologs) of a short film by test persons yielding different educational backgrounds and the verbal intelligence scores of the speakers. First, a one-way analysis of variance was performed to compare the monologs with the film transcription and to demonstrate that there are differences in the vocabulary used by the test persons yielding different verbal intelligence levels. Then, for the classification task, the monologs were represented as feature vectors using the classical TF–IDF weighting scheme. The Naive Bayes, k-nearest neighbors and Rocchio classifiers were tested. In this paper we describe and compare these classification approaches, define the optimal classification parameters and discuss the classification results obtained.

Language identification based on a discriminative text categorization technique

In this paper, we describe new results and improvements to a lan-guage identification (LID) system based on PPRLM previously introduced in [1] and [2]. In this case, we use as parallel phone recognizers the ones provided by the Brno University of Technology for Czech, Hungarian, and Russian lan-guages, and instead of using traditional n-gram language models we use a lan-guage model that is created using a ranking with the most frequent and discrim-inative n-grams. In this language model approach, the distance between the ranking for the input sentence and the ranking for each language is computed, based on the difference in relative positions for each n-gram. This approach is able to model reliably longer span information than in traditional language models obtaining more reliable estimations. We also describe the modifications that we have being introducing along the time to the original ranking technique, e.g., different discriminative formulas to establish the ranking, variations of the template size, the suppression of repeated consecutive phones, and a new clus-tering technique for the ranking scores. Results show that this technique pro-vides a 12.9% relative improvement over PPRLM. Finally, we also describe re-sults where the traditional PPRLM and our ranking technique are combined.

Using Kullback-Leibler Distance for Text Categorization

International audience

New kernel functions and learning methods for text and data mining

Recent advances in machine learning methods enable increasingly the automatic construction of various types of computer assisted methods that have been difficult or laborious to program by human experts. The tasks for which this kind of tools are needed arise in many areas, here especially in the fields of bioinformatics and natural language processing. The machine learning methods may not work satisfactorily if they are not appropriately tailored to the task in question. However, their learning performance can often be improved by taking advantage of deeper insight of the application domain or the learning problem at hand. This thesis considers developing kernel-based learning algorithms incorporating this kind of prior knowledge of the task in question in an advantageous way. Moreover, computationally efficient algorithms for training the learning machines for specific tasks are presented. In the context of kernel-based learning methods, the incorporation of prior knowledge is often done by designing appropriate kernel functions. Another well-known way is to develop cost functions that fit to the task under consideration. For disambiguation tasks in natural language, we develop kernel functions that take account of the positional information and the mutual similarities of words. It is shown that the use of this information significantly improves the disambiguation performance of the learning machine. Further, we design a new cost function that is better suitable for the task of information retrieval and for more general ranking problems than the cost functions designed for regression and classification. We also consider other applications of the kernel-based learning algorithms such as text categorization, and pattern recognition in differential display. We develop computationally efficient algorithms for training the considered learning machines with the proposed kernel functions. We also design a fast cross-validation algorithm for regularized least-squares type of learning algorithm. Further, an efficient version of the regularized least-squares algorithm that can be used together with the new cost function for preference learning and ranking tasks is proposed. In summary, we demonstrate that the incorporation of prior knowledge is possible and beneficial, and novel advanced kernels and cost functions can be used in algorithms efficiently.

Learning Methods and Algorithms for Semantic Text Classification across Multiple Domains

Information is nowadays a key resource: machine learning and data mining techniques have been developed to extract high-level information from great amounts of data. As most data comes in form of unstructured text in natural languages, research on text mining is currently very active and dealing with practical problems. Among these, text categorization deals with the automatic organization of large quantities of documents in priorly defined taxonomies of topic categories, possibly arranged in large hierarchies. In commonly proposed machine learning approaches, classifiers are automatically trained from pre-labeled documents: they can perform very accurate classification, but often require a consistent training set and notable computational effort. Methods for cross-domain text categorization have been proposed, allowing to leverage a set of labeled documents of one domain to classify those of another one. Most methods use advanced statistical techniques, usually involving tuning of parameters. A first contribution presented here is a method based on nearest centroid classification, where profiles of categories are generated from the known domain and then iteratively adapted to the unknown one. Despite being conceptually simple and having easily tuned parameters, this method achieves state-of-the-art accuracy in most benchmark datasets with fast running times. A second, deeper contribution involves the design of a domain-independent model to distinguish the degree and type of relatedness between arbitrary documents and topics, inferred from the different types of semantic relationships between respective representative words, identified by specific search algorithms. The application of this model is tested on both flat and hierarchical text categorization, where it potentially allows the efficient addition of new categories during classification. Results show that classification accuracy still requires improvements, but models generated from one domain are shown to be effectively able to be reused in a different one.

Use of Medical Subject Headings (MeSH) in Portuguese for categorizing web-based healthcare content

Introduction: Internet users are increasingly using the worldwide web to search for information relating to their health. This situation makes it necessary to create specialized tools capable of supporting users in their searches. Objective: To apply and compare strategies that were developed to investigate the use of the Portuguese version of Medical Subject Headings (MeSH) for constructing an automated classifier for Brazilian Portuguese-language web-based content within or outside of the field of healthcare, focusing on the lay public. Methods: 3658 Brazilian web pages were used to train the classifier and 606 Brazilian web pages were used to validate it. The strategies proposed were constructed using content-based vector methods for text classification, such that Naive Bayes was used for the task of classifying vector patterns with characteristics obtained through the proposed strategies. Results: A strategy named InDeCS was developed specifically to adapt MeSH for the problem that was put forward. This approach achieved better accuracy for this pattern classification task (0.94 sensitivity, specificity and area under the ROC curve). Conclusions: Because of the significant results achieved by InDeCS, this tool has been successfully applied to the Brazilian healthcare search portal known as Busca Saude. Furthermore, it could be shown that MeSH presents important results when used for the task of classifying web-based content focusing on the lay public. It was also possible to show from this study that MeSH was able to map out mutable non-deterministic characteristics of the web. (c) 2010 Elsevier Inc. All rights reserved.

Learning techniques for automatic email message tagging

A organização automática de mensagens de correio electrónico é um desafio actual na área da aprendizagem automática. O número excessivo de mensagens afecta cada vez mais utilizadores, especialmente os que usam o correio electrónico como ferramenta de comunicação e trabalho. Esta tese aborda o problema da organização automática de mensagens de correio electrónico propondo uma solução que tem como objectivo a etiquetagem automática de mensagens. A etiquetagem automática é feita com recurso às pastas de correio electrónico anteriormente criadas pelos utilizadores, tratando-as como etiquetas, e à sugestão de múltiplas etiquetas para cada mensagem (top-N). São estudadas várias técnicas de aprendizagem e os vários campos que compõe uma mensagem de correio electrónico são analisados de forma a determinar a sua adequação como elementos de classificação. O foco deste trabalho recai sobre os campos textuais (o assunto e o corpo das mensagens), estudando-se diferentes formas de representação, selecção de características e algoritmos de classificação. É ainda efectuada a avaliação dos campos de participantes através de algoritmos de classificação que os representam usando o modelo vectorial ou como um grafo. Os vários campos são combinados para classificação utilizando a técnica de combinação de classificadores Votação por Maioria. Os testes são efectuados com um subconjunto de mensagens de correio electrónico da Enron e um conjunto de dados privados disponibilizados pelo Institute for Systems and Technologies of Information, Control and Communication (INSTICC). Estes conjuntos são analisados de forma a perceber as características dos dados. A avaliação do sistema é realizada através da percentagem de acerto dos classificadores. Os resultados obtidos apresentam melhorias significativas em comparação com os trabalhos relacionados.

Automatic email organization

Introdução Actualmente, as mensagens electrónicas são consideradas um importante meio de comunicação. As mensagens electrónicas – vulgarmente conhecidas como emails – são utilizadas fácil e frequentemente para enviar e receber o mais variado tipo de informação. O seu uso tem diversos fins gerando diariamente um grande número de mensagens e, consequentemente um enorme volume de informação. Este grande volume de informação requer uma constante manipulação das mensagens de forma a manter o conjunto organizado. Tipicamente esta manipulação consiste em organizar as mensagens numa taxonomia. A taxonomia adoptada reflecte os interesses e as preferências particulares do utilizador. Motivação A organização manual de emails é uma actividade morosa e que consome tempo. A optimização deste processo através da implementação de um método automático, tende a melhorar a satisfação do utilizador. Cada vez mais existe a necessidade de encontrar novas soluções para a manipulação de conteúdo digital poupando esforços e custos ao utilizador; esta necessidade, concretamente no âmbito da manipulação de emails, motivou a realização deste trabalho. Hipótese O objectivo principal deste projecto consiste em permitir a organização ad-hoc de emails com um esforço reduzido por parte do utilizador. A metodologia proposta visa organizar os emails num conjunto de categorias, disjuntas, que reflectem as preferências do utilizador. A principal finalidade deste processo é produzir uma organização onde as mensagens sejam classificadas em classes apropriadas requerendo o mínimo número esforço possível por parte do utilizador. Para alcançar os objectivos estipulados, este projecto recorre a técnicas de mineração de texto, em especial categorização automática de texto, e aprendizagem activa. Para reduzir a necessidade de inquirir o utilizador – para etiquetar exemplos de acordo com as categorias desejadas – foi utilizado o algoritmo d-confidence. Processo de organização automática de emails O processo de organizar automaticamente emails é desenvolvido em três fases distintas: indexação, classificação e avaliação. Na primeira fase, fase de indexação, os emails passam por um processo transformativo de limpeza que visa essencialmente gerar uma representação dos emails adequada ao processamento automático. A segunda fase é a fase de classificação. Esta fase recorre ao conjunto de dados resultantes da fase anterior para produzir um modelo de classificação, aplicando-o posteriormente a novos emails. Partindo de uma matriz onde são representados emails, termos e os seus respectivos pesos, e um conjunto de exemplos classificados manualmente, um classificador é gerado a partir de um processo de aprendizagem. O classificador obtido é então aplicado ao conjunto de emails e a classificação de todos os emails é alcançada. O processo de classificação é feito com base num classificador de máquinas de vectores de suporte recorrendo ao algoritmo de aprendizagem activa d-confidence. O algoritmo d-confidence tem como objectivo propor ao utilizador os exemplos mais significativos para etiquetagem. Ao identificar os emails com informação mais relevante para o processo de aprendizagem, diminui-se o número de iterações e consequentemente o esforço exigido por parte dos utilizadores. A terceira e última fase é a fase de avaliação. Nesta fase a performance do processo de classificação e a eficiência do algoritmo d-confidence são avaliadas. O método de avaliação adoptado é o método de validação cruzada denominado 10-fold cross validation. Conclusões O processo de organização automática de emails foi desenvolvido com sucesso, a performance do classificador gerado e do algoritmo d-confidence foi relativamente boa. Em média as categorias apresentam taxas de erro relativamente baixas, a não ser as classes mais genéricas. O esforço exigido pelo utilizador foi reduzido, já que com a utilização do algoritmo d-confidence obteve-se uma taxa de erro próxima do valor final, mesmo com um número de casos etiquetados abaixo daquele que é requerido por um método supervisionado. É importante salientar, que além do processo automático de organização de emails, este projecto foi uma excelente oportunidade para adquirir conhecimento consistente sobre mineração de texto e sobre os processos de classificação automática e recuperação de informação. O estudo de áreas tão interessantes despertou novos interesses que consistem em verdadeiros desafios futuros.

Classificação e agregação automática de notícias desportivas

Mestrado em Engenharia Informática - Área de Especialização em Arquiteturas, Sistemas e Redes

Intelligent Internet Data Searching

In this thesis author approaches the problem of automated text classification, which is one of basic tasks for building Intelligent Internet Search Agent. The work discusses various approaches to solving sub-problems of automated text classification, such as feature extraction and machine learning on text sources. Author also describes her own multiword approach to feature extraction and pres-ents the results of testing this approach using linear discriminant analysis based classifier, and classifier combining unsupervised learning for etalon extraction with supervised learning using common backpropagation algorithm for multilevel perceptron.

Fitness assessment of document model

Document classification is a supervised machine learning process, where predefined category labels are assigned to documents based on the hypothesis derived from training set of labelled documents. Documents cannot be directly interpreted by a computer system unless they have been modelled as a collection of computable features. Rogati and Yang [M. Rogati and Y. Yang, Resource selection for domain-specific cross-lingual IR, in SIGIR 2004: Proceedings of the 27th annual international conference on Research and Development in Information Retrieval, ACM Press, Sheffied: United Kingdom, pp. 154-161.] pointed out that the effectiveness of document classification system may vary in different domains. This implies that the quality of document model contributes to the effectiveness of document classification. Conventionally, model evaluation is accomplished by comparing the effectiveness scores of classifiers on model candidates. However, this kind of evaluation methods may encounter either under-fitting or over-fitting problems, because the effectiveness scores are restricted by the learning capacities of classifiers. We propose a model fitness evaluation method to determine whether a model is sufficient to distinguish positive and negative instances while still competent to provide satisfactory effectiveness with a small feature subset. Our experiments demonstrated how the fitness of models are assessed. The results of our work contribute to the researches of feature selection, dimensionality reduction and document classification.

Multi-topic information filtering with a single user profile

In Information Filtering (IF) a user may be interested in several topics in parallel. But IF systems have been built on representational models derived from Information Retrieval and Text Categorization, which assume independence between terms. The linearity of these models results in user profiles that can only represent one topic of interest. We present a methodology that takes into account term dependencies to construct a single profile representation for multiple topics, in the form of a hierarchical term network. We also introduce a series of non-linear functions for evaluating documents against the profile. Initial experiments produced positive results.

D6.3 – Semantic Content Annotation Support

The Semantic Annotation component is a software application that provides support for automated text classification, a process grounded in a cohesion-centered representation of discourse that facilitates topic extraction. The component enables the semantic meta-annotation of text resources, including automated classification, thus facilitating information retrieval within the RAGE ecosystem. It is available in the ReaderBench framework (http://readerbench.com/) which integrates advanced Natural Language Processing (NLP) techniques. The component makes use of Cohesion Network Analysis (CNA) in order to ensure an in-depth representation of discourse, useful for mining keywords and performing automated text categorization. Our component automatically classifies documents into the categories provided by the ACM Computing Classification System (http://dl.acm.org/ccs_flat.cfm), but also into the categories from a high level serious games categorization provisionally developed by RAGE. English and French languages are already covered by the provided web service, whereas the entire framework can be extended in order to support additional languages.