Aircraft interior failure pattern recognition utilizing text mining and neural networks

Conselho Nacional de Desenvolvimento Científico e Tecnológico (CNPq)

Discovery driven analysis on semi-structured text data

Discovery Driven Analysis (DDA) is a common feature of OLAP technology to analyze structured data. In essence, DDA helps analysts to discover anomalous data by highlighting 'unexpected' values in the OLAP cube. By giving indications to the analyst on what dimensions to explore, DDA speeds up the process of discovering anomalies and their causes. However, Discovery Driven Analysis (and OLAP in general) is only applicable on structured data, such as records in databases. We propose a system to extend DDA technology to semi-structured text documents, that is, text documents with a few structured data. Our system pipeline consists of two stages: first, the text part of each document is structured around user specified dimensions, using semi-PLSA algorithm; then, we adapt DDA to these fully structured documents, thus enabling DDA on text documents. We present some applications of this system in OLAP analysis and show how scalability issues are solved. Results show that our system can handle reasonable datasets of documents, in real time, without any need for pre-computation.

Matching relevance features with ontological concepts

In order to comprehend user information needs by concepts, this paper introduces a novel method to match relevance features with ontological concepts. The method first discovers relevance features from user local instances. Then, a concept matching approach is developed for matching these features to accurate concepts in a global knowledge base. This approach is significant for the transition of informative descriptor and conceptional descriptor. The proposed method is elaborately evaluated by comparing against three information gathering baseline models. The experimental results shows the matching approach is successful and achieves a series of remarkable improvements on search effectiveness.

Link discovery for Chinese/English cross-language web information retrieval

Nowadays people heavily rely on the Internet for information and knowledge. Wikipedia is an online multilingual encyclopaedia that contains a very large number of detailed articles covering most written languages. It is often considered to be a treasury of human knowledge. It includes extensive hypertext links between documents of the same language for easy navigation. However, the pages in different languages are rarely cross-linked except for direct equivalent pages on the same subject in different languages. This could pose serious difficulties to users seeking information or knowledge from different lingual sources, or where there is no equivalent page in one language or another. In this thesis, a new information retrieval task—cross-lingual link discovery (CLLD) is proposed to tackle the problem of the lack of cross-lingual anchored links in a knowledge base such as Wikipedia. In contrast to traditional information retrieval tasks, cross language link discovery algorithms actively recommend a set of meaningful anchors in a source document and establish links to documents in an alternative language. In other words, cross-lingual link discovery is a way of automatically finding hypertext links between documents in different languages, which is particularly helpful for knowledge discovery in different language domains. This study is specifically focused on Chinese / English link discovery (C/ELD). Chinese / English link discovery is a special case of cross-lingual link discovery task. It involves tasks including natural language processing (NLP), cross-lingual information retrieval (CLIR) and cross-lingual link discovery. To justify the effectiveness of CLLD, a standard evaluation framework is also proposed. The evaluation framework includes topics, document collections, a gold standard dataset, evaluation metrics, and toolkits for run pooling, link assessment and system evaluation. With the evaluation framework, performance of CLLD approaches and systems can be quantified. This thesis contributes to the research on natural language processing and cross-lingual information retrieval in CLLD: 1) a new simple, but effective Chinese segmentation method, n-gram mutual information, is presented for determining the boundaries of Chinese text; 2) a voting mechanism of name entity translation is demonstrated for achieving a high precision of English / Chinese machine translation; 3) a link mining approach that mines the existing link structure for anchor probabilities achieves encouraging results in suggesting cross-lingual Chinese / English links in Wikipedia. This approach was examined in the experiments for better, automatic generation of cross-lingual links that were carried out as part of the study. The overall major contribution of this thesis is the provision of a standard evaluation framework for cross-lingual link discovery research. It is important in CLLD evaluation to have this framework which helps in benchmarking the performance of various CLLD systems and in identifying good CLLD realisation approaches. The evaluation methods and the evaluation framework described in this thesis have been utilised to quantify the system performance in the NTCIR-9 Crosslink task which is the first information retrieval track of this kind.

Text mining with semantic annotation : using enriched text representation for entity-oriented retrieval, semantic relation identification and text clustering

This project is a step forward in the study of text mining where enhanced text representation with semantic information plays a significant role. It develops effective methods of entity-oriented retrieval, semantic relation identification and text clustering utilizing semantically annotated data. These methods are based on enriched text representation generated by introducing semantic information extracted from Wikipedia into the input text data. The proposed methods are evaluated against several start-of-art benchmarking methods on real-life data-sets. In particular, this thesis improves the performance of entity-oriented retrieval, identifies different lexical forms for an entity relation and handles clustering documents with multiple feature spaces.

Topical pattern based document modelling and relevance ranking

For traditional information filtering (IF) models, it is often assumed that the documents in one collection are only related to one topic. However, in reality users’ interests can be diverse and the documents in the collection often involve multiple topics. Topic modelling was proposed to generate statistical models to represent multiple topics in a collection of documents, but in a topic model, topics are represented by distributions over words which are limited to distinctively represent the semantics of topics. Patterns are always thought to be more discriminative than single terms and are able to reveal the inner relations between words. This paper proposes a novel information filtering model, Significant matched Pattern-based Topic Model (SPBTM). The SPBTM represents user information needs in terms of multiple topics and each topic is represented by patterns. More importantly, the patterns are organized into groups based on their statistical and taxonomic features, from which the more representative patterns, called Significant Matched Patterns, can be identified and used to estimate the document relevance. Experiments on benchmark data sets demonstrate that the SPBTM significantly outperforms the state-of-the-art models.

Self-Organizing Information Fusion and Hierarchical Knowledge Discovery: A New Framework Using Artmap Neural Networks

Classifying novel terrain or objects from sparse, complex data may require the resolution of conflicting information from sensors woring at different times, locations, and scales, and from sources with different goals and situations. Information fusion methods can help resolve inconsistencies, as when eveidence variously suggests that and object's class is car, truck, or airplane. The methods described her address a complementary problem, supposing that information from sensors and experts is reliable though inconsistent, as when evidence suggests that an object's class is car, vehicle, and man-made. Underlying relationships among classes are assumed to be unknown to the autonomated system or the human user. The ARTMAP information fusion system uses distributed code representations that exploit the neural network's capacity for one-to-many learning in order to produce self-organizing expert systems that discover hierachical knowlege structures. The fusion system infers multi-level relationships among groups of output classes, without any supervised labeling of these relationships. The procedure is illustrated with two image examples, but is not limited to image domain.

Information sources selection methodology for recommender systems based on intrinsic characteristics and trust measure

El treball desenvolupat en aquesta tesi presenta un profund estudi i proveïx solucions innovadores en el camp dels sistemes recomanadors. Els mètodes que usen aquests sistemes per a realitzar les recomanacions, mètodes com el Filtrat Basat en Continguts (FBC), el Filtrat Col·laboratiu (FC) i el Filtrat Basat en Coneixement (FBC), requereixen informació dels usuaris per a predir les preferències per certs productes. Aquesta informació pot ser demogràfica (Gènere, edat, adreça, etc), o avaluacions donades sobre algun producte que van comprar en el passat o informació sobre els seus interessos. Existeixen dues formes d'obtenir aquesta informació: els usuaris ofereixen explícitament aquesta informació o el sistema pot adquirir la informació implícita disponible en les transaccions o historial de recerca dels usuaris. Per exemple, el sistema recomanador de pel·lícules MovieLens (http://movielens.umn.edu/login) demana als usuaris que avaluïn almenys 15 pel·lícules dintre d'una escala de * a * * * * * (horrible, ...., ha de ser vista). El sistema genera recomanacions sobre la base d'aquestes avaluacions. Quan els usuaris no estan registrat en el sistema i aquest no té informació d'ells, alguns sistemes realitzen les recomanacions tenint en compte l'historial de navegació. Amazon.com (http://www.amazon.com) realitza les recomanacions tenint en compte les recerques que un usuari a fet o recomana el producte més venut. No obstant això, aquests sistemes pateixen de certa falta d'informació. Aquest problema és generalment resolt amb l'adquisició d'informació addicional, se li pregunta als usuaris sobre els seus interessos o es cerca aquesta informació en fonts addicionals. La solució proposada en aquesta tesi és buscar aquesta informació en diverses fonts, específicament aquelles que contenen informació implícita sobre les preferències dels usuaris. Aquestes fonts poden ser estructurades com les bases de dades amb informació de compres o poden ser no estructurades com les pàgines web on els usuaris deixen la seva opinió sobre algun producte que van comprar o posseïxen. Nosaltres trobem tres problemes fonamentals per a aconseguir aquest objectiu: 1 . La identificació de fonts amb informació idònia per als sistemes recomanadors. 2 . La definició de criteris que permetin la comparança i selecció de les fonts més idònies. 3 . La recuperació d'informació de fonts no estructurades. En aquest sentit, en la tesi proposada s'ha desenvolupat: 1 . Una metodologia que permet la identificació i selecció de les fonts més idònies. Criteris basats en les característiques de les fonts i una mesura de confiança han estat utilitzats per a resoldre el problema de la identificació i selecció de les fonts. 2 . Un mecanisme per a recuperar la informació no estructurada dels usuaris disponible en la web. Tècniques de Text Mining i ontologies s'han utilitzat per a extreure informació i estructurar-la apropiadament perquè la utilitzin els recomanadors. Les contribucions del treball desenvolupat en aquesta tesi doctoral són: 1. Definició d'un conjunt de característiques per a classificar fonts rellevants per als sistemes recomanadors 2. Desenvolupament d'una mesura de rellevància de les fonts calculada sobre la base de les característiques definides 3. Aplicació d'una mesura de confiança per a obtenir les fonts més fiables. La confiança es definida des de la perspectiva de millora de la recomanació, una font fiable és aquella que permet millorar les recomanacions. 4. Desenvolupament d'un algorisme per a seleccionar, des d'un conjunt de fonts possibles, les més rellevants i fiable utilitzant les mitjanes esmentades en els punts previs. 5. Definició d'una ontologia per a estructurar la informació sobre les preferències dels usuaris que estan disponibles en Internet. 6. Creació d'un procés de mapatge que extreu automàticament informació de les preferències dels usuaris disponibles en la web i posa aquesta informació dintre de l'ontologia. Aquestes contribucions permeten aconseguir dos objectius importants: 1 . Millorament de les recomanacions usant fonts d'informació alternatives que sigui rellevants i fiables. 2 . Obtenir informació implícita dels usuaris disponible en Internet.

Automatic pattern-taxonomy extraction for web mining

In this paper, we propose a model for discovering frequent sequential patterns, phrases, which can be used as profile descriptors of documents. It is indubitable that we can obtain numerous phrases using data mining algorithms. However, it is difficult to use these phrases effectively for answering what users want. Therefore, we present a pattern taxonomy extraction model which performs the task of extracting descriptive frequent sequential patterns by pruning the meaningless ones. The model then is extended and tested by applying it to the information filtering system. The results of the experiment show that pattern-based methods outperform the keyword-based methods. The results also indicate that removal of meaningless patterns not only reduces the cost of computation but also improves the effectiveness of the system.

Reducing performance bias by intrinsically insensitive learning for unbalanced text mining

This thesis proposes three effective strategies to solve the significant performance-bias problem in imbalance text mining: (1) creation of a novel inexact field learning algorithm to overcome the dual-imbalance problem; (2) introduction of the one-class classification-framework to optimize classifier-parameters, and (3) proposal of a maximal-frequent-item-set discovery approach to achieve higher accuracy and efficiency.

Supporting personalised content management in smart health information portals

Information portals are seen as an appropriate platform for personalised healthcare and wellbeing information provision. Efficient content management is a core capability of a successful smart health information portal (SHIP) and domain expertise is a vital input to content management when it comes to matching user profiles with the appropriate resources. The rate of generation of new health-related content far exceeds the numbers that can be manually examined by domain experts for relevance to a specific topic and audience. In this paper we investigate automated content discovery as a plausible solution to this shortcoming that capitalises on the existing database of expert-endorsed content as an implicit store of knowledge to guide such a solution. We propose a novel content discovery technique based on a text analytics approach that utilises an existing content repository to acquire new and relevant content. We also highlight the contribution of this technique towards realisation of smart content management for SHIPs.

Granule-based knowledge representation for intra and inter transaction association mining

Abstract With the phenomenal growth of electronic data and information, there are many demands for the development of efficient and effective systems (tools) to perform the issue of data mining tasks on multidimensional databases. Association rules describe associations between items in the same transactions (intra) or in different transactions (inter). Association mining attempts to find interesting or useful association rules in databases: this is the crucial issue for the application of data mining in the real world. Association mining can be used in many application areas, such as the discovery of associations between customers’ locations and shopping behaviours in market basket analysis. Association mining includes two phases. The first phase, called pattern mining, is the discovery of frequent patterns. The second phase, called rule generation, is the discovery of interesting and useful association rules in the discovered patterns. The first phase, however, often takes a long time to find all frequent patterns; these also include much noise. The second phase is also a time consuming activity that can generate many redundant rules. To improve the quality of association mining in databases, this thesis provides an alternative technique, granule-based association mining, for knowledge discovery in databases, where a granule refers to a predicate that describes common features of a group of transactions. The new technique first transfers transaction databases into basic decision tables, then uses multi-tier structures to integrate pattern mining and rule generation in one phase for both intra and inter transaction association rule mining. To evaluate the proposed new technique, this research defines the concept of meaningless rules by considering the co-relations between data-dimensions for intratransaction-association rule mining. It also uses precision to evaluate the effectiveness of intertransaction association rules. The experimental results show that the proposed technique is promising.

Improvement to Chinese information retrieval by incorporating word segmentation and query expansion

The increasing diversity of the Internet has created a vast number of multilingual resources on the Web. A huge number of these documents are written in various languages other than English. Consequently, the demand for searching in non-English languages is growing exponentially. It is desirable that a search engine can search for information over collections of documents in other languages. This research investigates the techniques for developing high-quality Chinese information retrieval systems. A distinctive feature of Chinese text is that a Chinese document is a sequence of Chinese characters with no space or boundary between Chinese words. This feature makes Chinese information retrieval more difficult since a retrieved document which contains the query term as a sequence of Chinese characters may not be really relevant to the query since the query term (as a sequence Chinese characters) may not be a valid Chinese word in that documents. On the other hand, a document that is actually relevant may not be retrieved because it does not contain the query sequence but contains other relevant words. In this research, we propose two approaches to deal with the problems. In the first approach, we propose a hybrid Chinese information retrieval model by incorporating word-based techniques with the traditional character-based techniques. The aim of this approach is to investigate the influence of Chinese segmentation on the performance of Chinese information retrieval. Two ranking methods are proposed to rank retrieved documents based on the relevancy to the query calculated by combining character-based ranking and word-based ranking. Our experimental results show that Chinese segmentation can improve the performance of Chinese information retrieval, but the improvement is not significant if it incorporates only Chinese segmentation with the traditional character-based approach. In the second approach, we propose a novel query expansion method which applies text mining techniques in order to find the most relevant words to extend the query. Unlike most existing query expansion methods, which generally select the highly frequent indexing terms from the retrieved documents to expand the query. In our approach, we utilize text mining techniques to find patterns from the retrieved documents that highly correlate with the query term and then use the relevant words in the patterns to expand the original query. This research project develops and implements a Chinese information retrieval system for evaluating the proposed approaches. There are two stages in the experiments. The first stage is to investigate if high accuracy segmentation can make an improvement to Chinese information retrieval. In the second stage, a text mining based query expansion approach is implemented and a further experiment has been done to compare its performance with the standard Rocchio approach with the proposed text mining based query expansion method. The NTCIR5 Chinese collections are used in the experiments. The experiment results show that by incorporating the text mining based query expansion with the hybrid model, significant improvement has been achieved in both precision and recall assessments.