Document representation

As the volume of information in internet is increasing staggeringly therefore it is required to develop new methods for document retrieval and ranking them according to their relevance value as per the user query. Information Retrieval helps in searching meaningful and relevant documents present on semantic web. The concept of semantic similarity is used in variety of fields like artificial intelligence, natural language processing, cognitive science, Biomedical Informatics etc. Even after many developments maintaining updated version of documents and ordering them as per user’s request is still a big challenge. Our approach uses natural language processing techniques for preprocessing of documents that includes standard representation of documents, removal of stop words, stemming etc. Then the keywords are extracted using a technique. The novel approach suggested in this paper not only dependent on the syntactic structure of the document but the semantic structure also. It includes...

The paper demonstrates how the Laboratory Research Framework fits into the holistic Cognitive Framework for IR. It first discusses the Laboratory Framework with emphasis on its underlying assumptions and known limitations. This is followed by a view of interaction and relevance phenomena associated with IR evaluation and central to the understanding of IR. The ensuing section outlines how interactive IR is viewed from a Cognitive Framework, and 'light' interactive IR experiments are suggested performed by drawing on the latter framework's contextual possibilities. These include independent variables drawn from a collection, matching principles in a retrieval system, and the searcher's situation and task context. The paper ends with concluding points of summarization of issues encountered.

Abstract: In this paper we compare the relevance of information obtained from" discriminative" media and from" non-discriminative" media. Discriminative media are the ones which accumulate and deliver information using a heuristic selection of it. This can be made by humans, or by artificial intelligent systems, exhibiting some form of" knowledge". Non-discriminative media just collect and return information without any distinction. This can also be made by humans or by artificial systems, but there is no" knowledge" involved in ...

Canonical Information Retrieval systems perform a ranked keyword search strategy: Given a user's one-off information need (query), a list of documents, ordered by relevance, is returned. The main limitation of that “one fits all” approach is that long-term user interests are neglected in the search process, implicitly assuming that they are completely independent of the user query. Actually, there are information access scenarios that cannot be solved through a straightforward matching of queries and documents, since other elements influence the ...

How does an information user perceive a document as relevant? The literature on relevance has identified numerous factors affecting such a judgment. Taking a cognitive approach, this study focuses on the criteria users employ in making relevance judgment beyond topicality. On the basis of Grice's theory of communication, we propose a five‐factor model of relevance: topicality, novelty, reliability, understandability, and scope. Data are collected from a semicontrolled survey and analyzed by following a psychometric procedure. Topicality and novelty are found to be the two essential relevance criteria. Understandability and reliability are also found to be significant, but scope is not. The theoretical and practical implications of this study are discussed.

Using a fuzzy-logic-based calculus of linguistically quantified propositions we present FQUERY III+. a new, more "human-friendly" and easier-to-use implementation of a querying scheme proposed originally by Kacprzyk and Zioikowski to handle imprecise queries including a linguistic quantifier as, e.g. "find all records in which most (almost all, much more than 75%. . . or any other linguistic quantifier) of the imporfanf attributes (out of a specified set) are as desired (e.g. equal to five. more than IO, large, more or less equal to IS. etc.). FQUERY III + is an "add-on" to Ashton-Tate's dBase III Plus.

Traditionally, keyphrases (or keywords) have been manually assigned to documents by their authors or by human indexers. This, however, has become impractical due to the massive growth of documents—particularly short articles (e.g. microblogs, abstracts, snippets)—on the Internet each day, thus creating a need for systems that automatically extract keyphrases from documents. Automatic keyphrase extraction methods have generally taken either supervised or unsupervised approaches. Supervised methods extract keyphrases by using a training document set, thus acquiring knowledge from a global collection of texts. Conversely, unsupervised methods extract keyphrases by determining their relevance in a single-document context, without prior learning. We present a hybrid keyphrase extraction method for short articles, HybridRank, which leverages the benefits of both approaches. Our system implements modified versions of the TextRank (Mihalcea and Tarau, 2004)—unsupervised—and KEA (Witten et a...

In this paper we describe a type of data fusion involving the combination of evidence derived from multiple document representations. Our aim is to investigate if a composite representation can improve the online detection of novel events in a stream of broadcast news stories. This classification process otherwise known as first story detection FSD (or in the Topic Detection and Tracking pilot study as online new event detection [1]), is one of three main classification tasks defined by the TDT initiative. Our composite document representation consists of a semantic representation (based on the lexical chains derived from a text) and a syntactic representation (using proper nouns). Using the TDT1 evaluation methodology, we evaluate a number of document representation combinations using these document classifiers.

In this paper, we explore the effects of data fusion on First Story Detection [1] in a broadcast news domain. The data fusion element of this experiment involves the combination of evidence derived from two distinct representations of document content in a single cluster run. Our composite document representation consists of a concept representation (based on the lexical chains derived from a text) and free text representation (using traditional keyword index terms). Using the TDT1 evaluation methodology we evaluate a number of document representation strategies and propose reasons why our data fusion experiment shows performance improvements in the TDT domain.

Information is commonly reflected in news articles. However, texts are unstructured and thus demanding to analyze automatically. To identify and capture the facts in a news story we propose a novel approach, which utilizes natural language engineering. A combination of selected linguistic and statistical criteria enables the identification of grammatical units such as noun, verb, adjective, and adverb phrases. In literature, these entities are presumed to carry the meaning and the information expressed in English texts. In our study, we focus on determining multi-word features in articles related to the monetary policy conducted by the central bank in the USA, FED. The features are composed as attribute-value pairs, where the attributes represent grammatical units, which quantify the major event characteristics. The corresponding values are conditional expressions, which vary over time as facts evolve. The final set is aggregated over the corpus by the application of heuristic and syntax-based rules. Financial experts contributed to the project by providing expertise for the document interpretation.

In this paper, we address the problem of selection collections. This is important for locating responses in digital libraries. The aim of methods, which deal with the area of information retrieval, is to reduce the amount of the exchanged messages by selecting the best servers from the beginning of the search. We propose a new function of selection based on statistics. Our function takes into account the relevance degree of documents in order to rank collections. We have implemented and compared our function with other methods. The experimentations have shown that our proposition is very competitive.

Many modern applications produce and process XML data, which is queried in its both structural and textual component. This is especially useful if we consider a casual user who looks for information in web-based database systems or intranets containing XML data, like online shops, airline reservations, digital libraries catalogues or any other, and does not expect an exact answer. Many websites are built from document-centric XML documents . A remarkable characteristic of such XML data collections is that they are mostly heterogeneous, i.e. they contain domainfocused data, possibly valid w.r.t. various DTDs or XML schemes. XML documents can come from various sources. These collections can be managed as XML databases [5] as well as collections, providing an approximate way for users to search their contents. To ensure such functionality, it is required to approach these collections with both database and information retrieval (IR) methods. Current XML query languages like XPath and XQuery are applicable rather for data-centric than for document-centric XML data. Moreover, XML schemes are often necessary for their use. In other words, the languages are not longer appropriate for searching in such environments because they can not cope with the diversity of data. Hence, a research of integration of database querying and IR in context of XML is undoubtedly interesting and promising trend. Despite of the fact that a variety of systems that support such methods have been proposed, conventional IR techniques [2], e.g. vector space model, can be employed only restrictedly. The reason for it is that two types of queries should be dealt with: content-only (CO) queries, i.e. the traditional ones in IR, and content-and-structure (CAS) queries. A number of techniques to extend the vector space model have been designed, e.g. [6], [7], [8], [9], [11], and [12]. A usual critique of the mentioned approaches is that they not sufficiently reflect the structure of XML documents. A more advanced, twophase evaluation schema is proposed in . First, a modified vector space model is employed to obtain similarity scores for the textual nodes of XML trees. Then, the scores are propagated upward in the XML-trees with a possible modification and possibly new scores of other nodes are generated. In we described a matrix model based on an extension of the vector space model for XML data. A document D in a collection of XML documents C is represented by a matrix D, whose each row vector w t associated with a term t contains the weights of t for each path occurring in C. A query Q considered also as an XML tree is expressed as a matrix Q. The matrix model proposes to evaluate the degree of similarity of D with regard to the Q as the correlation between the matrices D and Q.

Your Ms V m reforenca Our iUe Narre raterence The author has granted a non-L'auteur a accordé une Licence non exclusive licence allowing the exclusive permettant à la National Library of Canada to Bibliothèque nationale du Canada de reproduce, loan, distribute or sell reproduire, prêter, distribuer ou copies of this thesis in microform, vendre des copies de cette thèse sous paper or electronic formats. la forme de microfiche/film, de reproduction sur papier ou sur format électronique. The author retains ownership of the L'auteur conserve la propriété du copyright in this thesis. Neither the droit d'auteur qui protège cette thèse. thesis nor substantial extracts fiom it Ni la thèse ni des extraits substantiels may be printed or otherwise de celle-ci ne doivent être imprimés reproduced without the author's ou autrement reproduits sans son permission. autorisation.

Documents exist in different formats. When we have document images, in order to access some part, preferably all, of the information contained in that images, we have to deploy a document image analysis application. Document images can be mostly textual or mostly graphical. If, for a user, a task is to retrieve document images, relevant to a query from a set, we must use indexing techniques. The documents and the query are translated in a common representation. Using a dissimilarity measure (between the query and the document representations) and a method to speed-up the search process we may find documents that are from the user point of view relevant to his query. The semantic gap between a document representation and the user implicit representation can lead to unsatisfactory results. If we want to access objects from document images that are relevant to the document semantic we must enter in a document understanding cycle. Understanding document images is made in systems that are (usually) domain dependent, and that are not applicable in general cases (textual and graphical document classes). In this paper we present a method to describe and then to index document images using frequently occurences of items. The intuition is that frequent items represents symbols in a certain domain and this document description can be related to the domain knowledge (in an unsupervised manner). The novelty of our method consists in using graph summaries as a description for document images. In our approach we use a bag (multiset) of graphs as description for document images. From the document images we extract a graph based representation. In these graphs, we apply graph mining techniques in order to find frequent and maximally subgraphs. For each document image we construct a bag with all frequent subgraphs found in the graph-based representation. This bag of "symbols" represents the description of the document.

This paper proposes a method to improve the existing methods to classify documents in to categories based on supervised machine learning technique. It includes converting the unstructured text data into a numerical vector form for representational learning. The key issue in text analysis and learning is to find an effective representational model. In case of large data sets the classical Bag of Words model outperforms the existing document representational learning methods in its simplicity in processing and accuracy. An improvement to this method using a fuzzy mapping model based on semantic similarity among words, beats the existing methods for classification. The method gives a comparatively simple representation and higher accuracy model. The results on real world data sets show an improved accuracy using the proposed method in supervised learning methods for classification.

Dans cet article, nous presentons une approche pour mesurer la similarite semantique entre des textes heterogenes et de qualite differente provenant de differentes sources Web. Notre approche commence par extraire le contenu des textes par deux methodes : (i) utilisation d'un systeme d'extraction que nous avons implante et qui identifie tous les mots contenus dans un texte donne, (ii) utilisation d'un thesaurus multilingue (AGROVOC). Ensuite, nous combinons les resultats des deux approches afin de mesurer la similarite entre les representations textuelles des documents. Afin d'evaluer les resultats, nous nous appuyons sur deux ensembles de donnees heterogenes issus du Web (tweets et articles scientifiques). (Resume d'auteur)

Risk Management is a practice composed by processes, methods and tools that allows managing risks in projects; this activity is typically started during the initial phase of a project and it continues during the whole project life cycle. Risk Identification implies to determine potential risks elements by using a consistent and structured method. Taxonomy-Based Risk Management involves using, during the Risk Identification tasks, a checklist of risk grouping structured according to different classes. This paper presents a new model to identify risks in projects based on the use of standard taxonomies. It is founded on experience and results feedback use. The proposed approach is adaptable to corporate and academic environments from different areas of knowledge and with diverse maturity levels.

Identifying topics and concepts associated with a set of documents is a critical task for information retrieval systems. One approach is to associate a query with a set of topics selected from a fixed ontology or vocabulary of terms. The core idea of this research is using Wikipedia articles and associated pages to make a topic ontology for this purpose. The benefit of this method is that Wikipedia is an online free-content encyclopedia which is developed through a social process and kept current by the Wikipedia community. In this paper the Persian Wikipedia has been analyzed in accordance to its articles and the category link graphs to extract a Persian pseudo-ontology. Thereafter, the created ontology has been applied through a query expansion algorithm to improve the performance of an information retrieval system. Our experiments show that it is possible to improve the precision of the information retrieval system by queries expansion based on Wikipedia.

The widespread use of the XML format for document representation and message exchange has influenced techniques for data integration in last years. A development of various XML languages, methods and tools gave rise to so called XML technology. Enterprise Information Integration (EII) requires an accurate, precise and complete understanding of the disparate data sources, the needs of the information consumers, and how these map to the business concepts of the enterprise. Any integration takes place in context of an Enterprise Information System (EIS). In the paper we explain various approaches to EII, its architectures as well as its association to Enterprise Application Integration (EAI). We introduce basic features and issues of EII and justify why XML technology contributes to finding sufficiently powerful support for tools for enabling EII. In particular, a database approach to XML provides a universal solution enabling to construct tools for achieving EII. In the paper we present some features of the XML technology, mainly its database part, and show how it is usable in EII.

Document classification presents difficult challenges due to the sparsity and the high dimensionality of text data, and to the complex semantics of the natural language. The traditional document representation is a word-based vector (Bag of Words, or BOW), where each dimension is associated with a term of the dictionary containing all the words that appear in the corpus. Although simple and commonly used, this representation has several limitations. It is essential to embed semantic information and conceptual patterns in order to enhance the prediction capabilities of classification algorithms. In this paper, we overcome the shortages of the BOW approach by embedding background knowledge derived from Wikipedia into a semantic kernel, which is then used to enrich the representation of documents. Our empirical evaluation with real data sets demonstrates that our approach successfully achieves improved classification accuracy with respect to the BOW technique, and to other recently developed methods.

Notre travail se situe dans le cadre d'un projet d'annotation descriptive, conceptuelle et thématique de corpus textuel. Dans le présent article, nous focalisons notre attention sur l'annotation conceptuelle et plus précisément sur la tâche d'extraction des termes pertinents. Pour ceci, nous proposons trois méthodes d'extraction de termes. Ces méthodes diffèrent du point de vue sources d'information utilisées (texte intégral, résumé et/ou annotations descriptives) et de point de vue méthodes d'estimation de la pertinence des termes (calculs basés sur la fréquence des termes, calculs basés sur les relations d'association entre les termes). Afin d'évaluer ces trois méthodes nous avons eu recours à un ensemble de 19 classifieurs. Une discussion sur nos résultats met en évidence l'importance de l'exploitation des réseaux de cooccurrences dans un processus d'extraction de termes notamment dans l'objectif d'enrichir l'indexation conceptuelle de documents textuels.

In this paper, we introduce TextRank-a graph-based ranking model for text processing, and show how this model can be successfully used in natural language applications. In particular, we propose two innovative unsupervised methods for keyword and sentence extraction, and show that the results obtained compare favorably with previously published results on established benchmarks.

In Automatic Text Processing tasks, documents are usually represented in the bag-ofwords space. However, this representation does not take into account the possible relations between words. We propose here a review of a family of document density estimation models for representing documents. Inside this family we derive another possible model: the Theme Topic Mixture Model (TTMM). This model assumes two types of relations among textual data. Topics link words to each other and Themes gather documents with particular distribution over the topics. An experiment reports the performance of the different models in this family over a common task.

In Automatic Text Processing tasks, documents are usually represented in the bag-ofwords space. However, this representation does not take into account the possible relations between words. We propose here a review of a family of document density estimation models for representing documents. Inside this family we derive another possible model: the Theme Topic Mixture Model (TTMM). This model assumes two types of relations among textual data. Topics link words to each other and Themes gather documents with particular distribution over the topics. An experiment reports the performance of the different models in this family over a common task.

New text analysis softwares issued from fields of research such as Machine Learning and Natural Languages Processing prove to be relevant tools for the language sciences. Littératron is a new data-processing tool for the automatic extraction of syntactic patterns, designed at LIP6 by Jean-Gabriel Ganascia. Associated with a linear text analyser, it reveals the stylistic peculiarities of a text. We will see that Littératron carries out a linguistic diagnosis of learners if used in language sciences, especially in the field of acquisition of written French as a foreign language. The learner can be from a heterogeneous group (various language levels and various mother tongues) or from a homogeneous group (only one language level and one mother tongue, here, Arabic). The interest of this approach is related to three fields: first, language didactics, on a purely educational basis; next, computational linguistics; finally, computer-assisted learning. De nouveaux logiciels d"analyse textuelle tirent partie des progrès récents effectués en apprentissage symbolique et dans le traitement automatique des langues naturelles. Conçu au LIP6 par Jean-Gabriel Ganascia, le Littératron est l'un d'entre eux ; il extrait automatiquement des motifs syntaxiques i à partir de textes écrits en langage naturel. Plus exactement, le Littératron prend comme entré un arbre d'analyse syntaxique et donne en sortie un certain nombre de motifs syntaxiques récurrents. Associé à un analyseur de textes, qui engendre l'arbre d'analyse syntaxique à partir de textes écrits en langage naturel, il révèle les singularités stylistiques de ces textes. Nous allons voir qu"utilisé en sciences du langage, dans le domaine de l"acquisition du français écrit, le Littératron permet d'effectuer un diagnostic linguistique de l"apprenant, que celui-ci provienne d"une classe de langue hétérogène (différentes langues maternelles) ou homogène (une seule langue maternelle, en l"occurrence ici l"arabe). L"intérêt de cette approche concerne trois domaines : d"une part la didactique des langues, à titre éducatif ; d"autre part, la linguistique computationnelle, et enfin l"enseignement assisté par ordinateur.

A massive amount of unstructured data, in this information age, is composed of document collections. Examples include news articles, blog posts, scholarly publications, and reports generated by organizations as well as people. Many data mining and machine learning algorithms have been developed in the past decade to support text mining in many applications. Text mining applications cover a wide range of tasks spanning from personal information management — to organizational decision making, to disease control through epidemic prediction, and to intelligence analysis for national security. One bottleneck has been dominating data mining and machine learning theories for all these text mining applications — the quality of the outcomes of the algorithms depends on the quality of the representation of the documents. My research exploits imagery and textual content of documents to create high quality representations for documents, document tokens (e.g., names of people), and image snippet...

Most text classification systems use bag-of-words representation of documents to find the classification target function. Linguistic structures such as morphology, syntax and semantic are completely neglected in the learning process. This paper proposes a new document representation that, while including its context independent sentence meaning, is able to be used by a structured kernel function, namely the direct product kernel. The proposal is evaluated using a dataset of articles from a Portuguese daily newspaper and classifiers are built using the SVM algorithm. The results show that this structured representation, while only partially describing document's significance has the same discriminative power over classes as the traditional bag-of-words approach.

Most studies on authorship identification reported a drop in the identification result when the number of authors exceeds 20-25. In this paper, we introduce a new user representation to address this problem and split classification across two layers. There are at least 3 novelties in this paper. First, the two-layer approach allows applying authorship identification over larger number of authors (tested over 100 authors), and it is extendable. The authors are divided into groups that contain smaller number of authors. Given an anonymous document, the primary layer detects the group to which the document belongs. Then, the secondary layer determines the particular author inside the selected group. In order to extract the groups linking similar authors, clustering is applied over users rather than documents. Hence, the second novelty of this paper is introducing a new user representation that is different from document representation. Without the proposed user representation, the clustering over documents will result in documents of author(s) distributed over several clusters, instead of a single cluster membership for each author. Third, the extracted clusters are descriptive and meaningful of their users as the dimensions have psychological backgrounds. For authorship identification, the documents are labelled with the extracted groups and fed into machine learning to build classification models that predicts the group and author of a given document. The results show that the documents are highly correlated with the extracted corresponding groups, and the proposed model can be accurately trained to determine the group and the author identity.

In metaphorical conceptualization, the structure of one conceptual system is projected onto another. Our previous work suggest that idiomaticity in specialized languages is based in part on this phenomenon. However, since metaphorical conceptualizations are cognitive in nature, they are not easy to identify and characterize. Hypothesizing that it is possible to describe these conceptualizations based on their expression in discourse, we propose a semantic corpus annotation method using XML. The system we have developed allows for the systematic study of metaphorical conceptualizations and constitutes an elegant and efficient means of extracting qualitative and quantitative data that could not otherwise be obtained. Résumé La conceptualisation métaphorique consiste à projeter la structure d'un système conceptuel sur une autre. Nos travaux antérieurs suggèrent que l'idiomaticité repose en grande partie sur ce phénomène, qui se manifeste, entre autres, dans les langues de spécialité. Toutefois, comme elle se situe au plan de la pensée, la conceptualisation métaphorique ne se laisse pas appréhender facilement. Faisant l'hypothèse qu'il est possible de la caractériser par l'intermédiaire de ses réalisations en discours, nous proposons une méthode d'annotation sémantique de corpus faisant appel au langage XML. Le système mis en oeuvre permet une systématisation de l'étude de la conceptualisation métaphorique et autorise, de façon élégante, l'obtention de données qualitatives et quantitatives autrement impossibles à obtenir.

Public reporting burden for this collection of information is estimated to average 1 hour per response, including the time for reviewing instructions, searching existing data sources, gathering and maintaining the data needed, and completing and reviewing the collection of information. Send comment regarding this burden estimate or any other aspect of this collection of information, including suggestions for reducing this burden, to Washington Headquarters Services. Directorate for information Operations and Reports,

One of the most challenging aspects of developing information systems is the processing and management of large volumes of information. One way to overcome this problem is to implement efficient data indexing and classification systems. As large volumes of generated data comprise of non-structured textual data, developing text processing, management and indexing frameworks can play an important role in providing users with accurate information according to their preferences. In this paper, a novel method of semantic information processing, management and indexing is introduced. The main goals of this study is to integrate structured knowledge of ontology and Knowledge Bases (KBs) in the core components of the method, to enrich the contents of the documents, to have multi-level semantic network representation of textual resources, to introduce a hybrid weighting schema (salient score) and finally to propose a hybrid method of semantic similarity computation. The structured knowledge of ontology and KBs are integrated from all aspects of the proposed method. The obtained results indicate the accuracy and optimal performance of the proposed framework. The obtained results suggest that using knowledge-based models leads to higher performance and accuracy in identifying and classifying documents according to user preferences; however, if learning-based models are not provided with sufficient amount of training data, they cannot yield satisfying results. The results also demonstrate that the complete integration of ontology and KBs in information systems can significantly contribute to a better representation of documents and evidently superior functionality of information processing, management and indexing systems.

In this research, we enhanced the performance of Support Vector Machine (SVM) in text classification by applying semantic-knowledge enrichment. We propose using semantic-knowledge enrichment scheme to inject new concepts into the original contents of the text documents. A pre-processing technique is proposed for cleaning and extracting features for generating semantic concepts through using WordNet database and the open source Natural Language Toolkit (NLTK). Additionally, the combined online variation Bayes algorithm and the Latent Dirichlet Allocation model are used as a dimensionality reduction technique to generate abstract concepts from the raw text. In our experiment, we clarified the process of preparing data for cleaning, transformation and weighting the features vectors in a multi-dimensional space as a step to measure the performance metrics of SVM, before and after applying our proposed approach on two different datasets. K-Fold Cross-Validation technique is used to validate our proposed approach. Moreover, a confusion matrix is implemented to measure the accuracy and macro-averages of precision, recall and f1 measurements. The result of the evaluation showed improvements in term of accuracy from 94% to 98.3% for the dataset-1, and from 88% to 93% for dataset-2. Moreover, the training time of the classifier in terms of seconds was reduced to 32% and 17% for dataset-1 and dataset-2 respectively, in comparison with the training time of the original data before applying our proposed enrichment scheme.

Text document clustering plays an important role in providing better document retrieval, document browsing, and text mining. Traditionally, clustering techniques do not consider the semantic relationships between words, such as synonymy and hypernymy. To exploit semantic relationships, ontologies such as WordNet have been used to improve clustering results. However, WordNet-based clustering methods mostly rely on single-term analysis of text; they do not perform any phrase-based analysis. In addition, these methods utilize synonymy to identify concepts and only explore hypernymy to calculate concept frequencies, without considering other semantic relationships such as hyponymy. To address these issues, we combine detection of noun phrases with the use of WordNet as background knowledge to explore better ways of representing documents semantically for clustering. First, based on noun phrases as well as single-term analysis, we exploit different document representation methods to analyze the effectiveness of hypernymy, hyponymy, holonymy, and meronymy. Second, we choose the most effective method and compare it with the WordNet-based clustering method proposed by others. The experimental results show the effectiveness of semantic relationships for clustering are (from highest to lowest): hypernymy, hyponymy, meronymy, and holonymy. Moreover, we found that noun phrase analysis improves the WordNet-based clustering method.

MAD (Movie Authoring and Design) is a novel design and authoring system that facilitates the process of creating dynamic visual presentations. MAD aids this process by simultaneously allowing easy creation and modification of structured motion pictures and visualization of the result of the modifications. The principles behind MAD include hierarchical multimedia document representation, the flexible inclusion and combination of words,

This paper presents the first results of a study on the bargaining process of web standards in World Wide Web Consortium (W3C) arenas. This process is analysed through bargaining habits and through networks of actors who take part in it. The data collection is based on infor mation about individuals who play the game: Who ar e they? In the name of w hom do they talk? etc. Of ficial chats and for ums, web home pages and per sonal pages are exploited by ad hoc craw ling methods, and stored in an XML data war ehous e. This war ehous e is based on the semi-structured database model, which is a flexible treepatterned schema, very different from the well-known classical relational tables and traditional data warehousing techniques. This structure can impr ove the management of sociological inquiries, by allowing an evolutive strategy of querying, based on categorization. It simplifies the research process prior to quantitative work (statistics, factorial analyses, structural network analyses). In this paper, we present both the foundations and architecture of the warehouse, and the sociological results that have been gathered so far using these novel techniques.

An overwhelming number of users use Ebooks as their primary formats. Gone are the days when buying a physical copy was the only thing to do. Even though technology has advanced to extreme extents in display and visualization of texts (display technology like elink, formats such as PDFs), our team is of the opinion that enough time and effort has not been put in into making data more audiovisual, and easier to understand and grasp. It is with his aim that we attacked this problem. Researching formats that are best suited for visualization and using NLP techniques to implement our end game, is the crux of this project.

Feature selection (FS) is a widely used method for removing redundant or irrelevant features to improve classification accuracy and decrease the model's computational cost. In this paper, we present an improved method (referred to hereafter as RARF) for Arabic text classification (ATC) that employs the term frequency-inverse document frequency (TF-IDF) and Word2Vec embedding technique to identify words that have a particular semantic relationship. In addition, we have compared our method with four benchmark FS methods namely principal component analysis (PCA), linear discriminant analysis (LDA), chi-square, and mutual information (MI). Support vector machine (SVM), k-nearest neighbors (K-NN), and naive Bayes (NB) are three machine learning based algorithms used in this work. Two different Arabic datasets are utilized to perform a comparative analysis of these algorithms. This paper also evaluates the efficiency of our method for ATC on the basis of performance metrics viz accuracy, precision, recall, and F-measure. Results revealed that the highest accuracy achieved for the SVM classifier applied to the Khaleej-2004 Arabic dataset with 94.75%, while the same classifier recorded an accuracy of 94.01% for the Watan-2004 Arabic dataset.

Feature selection (FS) is a widely used method for removing redundant or irrelevant features to improve classification accuracy and decrease the model's computational cost. In this paper, we present an improved method (referred to hereafter as RARF) for Arabic text classification (ATC) that employs the term frequency-inverse document frequency (TF-IDF) and Word2Vec embedding technique to identify words that have a particular semantic relationship. In addition, we have compared our method with four benchmark FS methods namely principal component analysis (PCA), linear discriminant analysis (LDA), chi-square, and mutual information (MI). Support vector machine (SVM), k-nearest neighbors (K-NN), and naive Bayes (NB) are three machine learning based algorithms used in this work. Two different Arabic datasets are utilized to perform a comparative analysis of these algorithms. This paper also evaluates the efficiency of our method for ATC on the basis of performance metrics viz accuracy, precision, recall, and F-measure. Results revealed that the highest accuracy achieved for the SVM classifier applied to the Khaleej-2004 Arabic dataset with 94.75%, while the same classifier recorded an accuracy of 94.01% for the Watan-2004 Arabic dataset.

Answering mobile users' queries intelligently is one of the significant challenges in information retrieval (IR) in intelligent systems. Current popular Quranic retrieval application ranks the document by counting the occurrences of each of the terms and ignoring any other information in the document to solve the Verses of Quran retrieval problem. Considering the proximity between the query terms assists in the efficiency of ranking results and increases IR performance. The Spectral-Based Information Retrieval Model (SBIRM) considers the query terms' proximity by examining the term patterns that occur in the documents. To do this, SBIRM utilizes term signal representation and discrete wavelet transform (DWT). In this paper, we solve the Verses of Quran retrieval problem by proposing a novel document model, termed the Dynamic Document Model with Discrete Wavelet Transforms (DDMDWT). The DDMDWT exploits the variations in Verses of Quran length and mathematical transforms for document representation. The proposed model will enhance the existing term signal concept by additionally taking into consideration differing lengths of Verses of Quran. We designed and implemented an intelligent Quranic retrieval (IQR) Android application. In this IQR, the DDMDWT model contributes to reducing the time complexity of SBIRM and decreasing the index size by 20.98%, all while achieving improvement in precision, recall, F-measure, and MAP with compared to SBIRM. This paper also demonstrates how the DDMDWT model delivers a notable increase in the precision of the P@1 and P@3.

Language resources are typically defined and created for application in speech technology contexts, but the documentation of languages which are unlikely ever to be provided with enabling technologies nevertheless plays an important role in defining the heritage of a speech community and in the provision of basic insights into the language oriented components of human cognition. This is particularly true of endangered languages. The present case study concerns the documentation both of the birth and of the endangerment within a rather short space of time of a 'spirit language', Medefaidrin, created and used as a vehicular language by a religious community in SouthEastern Nigeria. The documentation shows phonological, orthographic, morphological, syntactic and textual typological features of Medefaidrin which indicate that typological properties of English were a model for the creation of the language, rather than typological properties of the enclaving language, Ibibio. The documentation is designed as part of the West African Language Archive (WALA), following OLAC metadata standards.

This paper describes an algorithm for document representation in a reduced vectorial space by a process of feature extraction. The algorithm is applied and evaluated in the context of the supervised classification of news articles from the collection of Le Monde newspaper issued in the years 2003 and 2004. We are generating a document representation (or profile), in a space of 800 dimensions, represented by semantic tags from a machine-readable dictionary. We are dealing with two issues: the synonymy handled by thematic conflation and polysemy for which we have developed a statistical method for word-sense disambiguation. We propose four variants for the profile generation (of a document) depending on whether a recursive system is used or not, and whether a corrective factor for polysemous words is taken into account or not. To determine the best classifier provided by our algorithm we have evaluated 32 variants, depending on the algorithm type (as previously) and on three other parameters that influence the document representation: grammatical category selection, 15% reduction of the profile, and a stop-list of semantic tags. The evaluation is done on a set of documents from six categories by calculating the precision, the recall and the F-measure to determine the best algorithm related to the threshold detection. Some parameters (like profile reduction) have low influence on the classifier performance and others (corrective factor for the ambiguous words, stop-list) improve it noticeably. Résumé Cet article présente un algorithme de représentation vectorielle de textes qui réalise une réduction de l'espace de représentation par une méthode d'extraction d'attributs. Nous en avons évalué les performances dans le cadre de la classification automatique supervisée à partir d'un ensemble de documents textes issus de la collection du journal « Le Monde » des années 2003 et 2004. Notre algorithme génère une représentation de document (appelée « profil »), dans un espace d'environ 800 dimensions représentées par des étiquettes sémantiques issues d'un dictionnaire électronique. Parallèlement à cette réduction de l'espace de représentation, celui-ci traite également les problèmes de synonymie, gérés par regroupement thématique, et de polysémie, à l'aide d'une méthode statistique de désambiguïsation sémantique. Nous proposons quatre variations différentes pour générer un profil de document en fonction de l'utilisation ou non du système récursif, et de l'ajout d'un coefficient de pénalité corrigeant l'influence des mots polysémiques. Afin de déterminer le meilleur classifieur possible issu de notre algorithme, nous avons alors généré un ensemble de 32 classifieurs en étudiant en plus de ces quatre variations, l'influence de trois paramètres supplémentaires (sélection des catégories grammaticales, réduction de 15% du profil, application d'une stop-liste d'étiquettes sémantiques) agissant sur la représentation des documents. L'évaluation de la performance de ces classifieurs s'est faite sur la base de documents issus de six catégories. Le calcul du rappel, de la précision et de la mesure F 1 nous a permis de déterminer l'algorithme optimal en fonction du type de détection de seuil utilisé (dépendant de l'application envisagée). Certains paramètres (réduction du profil) ont ainsi une faible influence sur la performance du classifieur, tandis que d'autres (coefficient de correction pour les mots ambigus, stop-liste) améliorent sensiblement la performance.

This paper describes an algorithm for document representation in a reduced vectorial space by a process of feature extraction. The algorithm is evaluated in the context of the supervised classification of news articles. We are generating a document representation (profile) represented by semantic tags from a machine-readable dictionary. We are dealing with synonymy handled by thematic conflation, and polysemy for which we have developed a statistical method for word-sense disambiguation. We propose four variants for the profile generation depending on whether a recursive system is used or not, and whether a corrective factor for polysemous words is taken into account or not. We have evaluated 32 variants, depending on the algorithm type and on three other parameters: grammatical category selection, 15% reduction of the profile, and a stop-list of semantic tags. Some parameters (like profile reduction) have low influence on the classifier performance and others (corrective factor for the ambiguous words, stop-list) improve the performance noticeably.

Abstract-OntoGen is a semi-automatic and data-driven ontology editor focusing on editing of topic ontologies. It utilizes text mining tools to make the ontology-related tasks simpler to the user. This focus on building ontologies from textual data is what we are trying to bridge. We have successfully extended OntoGen to work with image data and allow for ontology construction and editing on collections of labeled or unlabeled images. Browsing large heterogenous image collections efficiently is certainly a challenging task -and we feel that semiautomatic ontology construction, as described in this paper, makes this task easier.

Documents are often marked up in XML-based tagsets to delineate major structural components such as headings, paragraphs, figure captions and so on, without much regard to their eventual displayed appearance. And yet these same abstract documents, after many transformations and 'typesetting' processes, often emerge in the popular format of Adobe PDF, either for dissemination or archiving. XML, PDF, document structure transformation.

Adobe's Acrobat software, released in June 1993, is based around a new Portable Document Format (PDF) which offers the possibility of being able to view and exchange electronic documents, independent of the originating software, across a wide variety of supported hardware platforms (PC, Macintosh, Sun UNIX etc.). The principal features of Acrobat are reviewed and its importance for libraries discussed in the context of experience already gained from the CAJUN project (CD-ROM Acrobat TM Journals Using Networks). This two-year project, funded by two well-known journal publishers, is investigating the use of Acrobat software for the 'electronic' dissemination of journals, on CD-ROM and over networks.

This paper draws a parallel between document preparation and the traditional processes of compilation and link editing for computer programs. A block-based document model is described which allows for separate compilation of various portions of a document. These portions are brought together and merged by a linker program, called dlink, whose pilot implementation is based on ditroff and on its underlying intermediate code. In the light of experiences with dlink the requirements for a universal 'object-module language' for documents are discussed. These requirements often resemble the characteristics of the intermediate codes used by programming-language compilers but with interesting extra constraints which arise from the way documents are 'executed'.

The two complementary de facto standards for the publication of electronic documents are HTML on the World Wide Web and Adobe's Acrobat viewers using PDF (Portable Document Format). A brief overview is given of these two systems followed by an analysis of why the embedded, and very concrete, nature of their hypertext links leads to great problems with keeping the 'hyperstructure' up to date. An SGML-based syntax is presented, adapted from that used in the Microcosm system, which allows links and other hypertextual material to be kept in an abstract form in separate link bases. The links can then be interpreted or compiled at any stage and applied, in the correct format to some specific representation such as HTML or PDF. This approach is of great value in keeping hyperlinks relevant, up-to-date and in a common link-base which is independent of the finally delivered electronic document format. Four models are discussed for allowing publishers to insert links into documents at a late stage, e.g at the time that the document is requested by the end-user from the publisher. These methods ensure that disseminated papers always contain up-to-date links. The techniques discussed have been implemented using a combination of Acrobat plug-ins, Web servers and Web browsers.