Document engineering approaches toward scalable and structured multimedia, web and printable documents

Document engineering is the computer science discipline that investigates systems for documents in any form and in all media. As with the relationship between software engineering and software, document engineering is concerned with principles, tools and processes that improve our ability to create, manage, and maintain documents (http://www.documentengineering.org). The ACM Symposium on Document Engineering is an annual meeting of researchers active in document engineering: it is sponsored by ACM by means of the ACM SIGWEB Special Interest Group. In this editorial, we first point to work carried out in the context of document engineering, which are directly related to multimedia tools and applications. We conclude with a summary of the papers presented in this special issue.

Structural patterns for document engineering: from an empirical bottom-up analysis to an ontological theory

This thesis aims at investigating a new approach to document analysis based on the idea of structural patterns in XML vocabularies. My work is founded on the belief that authors do naturally converge to a reasonable use of markup languages and that extreme, yet valid instances are rare and limited. Actual documents, therefore, may be used to derive classes of elements (patterns) persisting across documents and distilling the conceptualization of the documents and their components, and may give ground for automatic tools and services that rely on no background information (such as schemas) at all. The central part of my work consists in introducing from the ground up a formal theory of eight structural patterns (with three sub-patterns) that are able to express the logical organization of any XML document, and verifying their identifiability in a number of different vocabularies. This model is characterized by and validated against three main dimensions: terseness (i.e. the ability to represent the structure of a document with a small number of objects and composition rules), coverage (i.e. the ability to capture any possible situation in any document) and expressiveness (i.e. the ability to make explicit the semantics of structures, relations and dependencies). An algorithm for the automatic recognition of structural patterns is then presented, together with an evaluation of the results of a test performed on a set of more than 1100 documents from eight very different vocabularies. This language-independent analysis confirms the ability of patterns to capture and summarize the guidelines used by the authors in their everyday practice. Finally, I present some systems that work directly on the pattern-based representation of documents. The ability of these tools to cover very different situations and contexts confirms the effectiveness of the model.

The Use of Latent Semantic Indexing to Mitigate OCR Effects of Related Document Images

Due to both the widespread and multipurpose use of document images and the current availability of a high number of document images repositories, robust information retrieval mechanisms and systems have been increasingly demanded. This paper presents an approach to support the automatic generation of relationships among document images by exploiting Latent Semantic Indexing (LSI) and Optical Character Recognition (OCR). We developed the LinkDI (Linking of Document Images) service, which extracts and indexes document images content, computes its latent semantics, and defines relationships among images as hyperlinks. LinkDI was experimented with document images repositories, and its performance was evaluated by comparing the quality of the relationships created among textual documents as well as among their respective document images. Considering those same document images, we ran further experiments in order to compare the performance of LinkDI when it exploits or not the LSI technique. Experimental results showed that LSI can mitigate the effects of usual OCR misrecognition, which reinforces the feasibility of LinkDI relating OCR output with high degradation.

Evaluating Invariances in Document Layout Functions

With the development of variable-data-driven digital presses - where each document printed is potentially unique - there is a need for pre-press optimization to identify material that is invariant from document to document. In this way rasterisation can be confined solely to those areas which change between successive documents thereby alleviating a potential performance bottleneck. Given a template document specified in terms of layout functions, where actual data is bound at the last possible moment before printing, we look at deriving and exploiting the invariant properties of layout functions from their formal specifications. We propose future work on generic extraction of invariance from such properties for certain classes of layout functions.

Speculative document evaluation

Optimisation of real world Variable Data printing (VDP) documents is a dicult problem because the interdependencies between layout functions may drastically reduce the number of invariant blocks that can be factored out for pre-rasterisation. This paper examines how speculative evaluation at an early stage in a document-preparation pipeline, provides a generic and effective method of optimising VDP documents that contain such interdependencies. Speculative evaluation will be at its most effective in speeding up print runs if sets of layout invariances can either be discovered automatically, or designed into the document at an early stage. In either case the expertise of the layout designer needs to be supplemented by expertise in exploiting potential invariances and also in predicting the effects of speculative evaluation on the caches used at various stages in the print production pipeline.

Extracting reusable document components for variable data printing

Variable Data Printing (VDP) has brought new flexibility and dynamism to the printed page. Each printed instance of a specific class of document can now have different degrees of customized content within the document template. This flexibility comes at a cost. If every printed page is potentially different from all others it must be rasterized separately, which is a time-consuming process. Technologies such as PPML (Personalized Print Markup Language) attempt to address this problem by dividing the bitmapped page into components that can be cached at the raster level, thereby speeding up the generation of page instances. A large number of documents are stored in Page Description Languages at a higher level of abstraction than the bitmapped page. Much of this content could be reused within a VDP environment provided that separable document components can be identified and extracted. These components then need to be individually rasterisable so that each high-level component can be related to its low-level (bitmap) equivalent. Unfortunately, the unstructured nature of most Page Description Languages makes it difficult to extract content easily. This paper outlines the problems encountered in extracting component-based content from existing page description formats, such as PostScript, PDF and SVG, and how the differences between the formats affects the ease with which content can be extracted. The techniques are illustrated with reference to a tool called COG Extractor, which extracts content from PDF and SVG and prepares it for reuse.

L'ingénierie des documents d'affaires dans le cadre du web sémantique

Dans cette thèse, nous présentons les problèmes d’échange de documents d'affaires et proposons une méthode pour y remédier. Nous proposons une méthodologie pour adapter les standards d’affaires basés sur XML aux technologies du Web sémantique en utilisant la transformation des documents définis en DTD ou XML Schema vers une représentation ontologique en OWL 2. Ensuite, nous proposons une approche basée sur l'analyse formelle de concept pour regrouper les classes de l'ontologie partageant une certaine sémantique dans le but d'améliorer la qualité, la lisibilité et la représentation de l'ontologie. Enfin, nous proposons l’alignement d'ontologies pour déterminer les liens sémantiques entre les ontologies d'affaires hétérogènes générés par le processus de transformation pour aider les entreprises à communiquer fructueusement.

Combining Taxonomies using Word2vec

Taxonomies have gained a broad usage in a variety of fields due to their extensibility, as well as their use for classification and knowledge organization. Of particular interest is the digital document management domain in which their hierarchical structure can be effectively employed in order to organize documents into content-specific categories. Common or standard taxonomies (e.g., the ACM Computing Classification System) contain concepts that are too general for conceptualizing specific knowledge domains. In this paper we introduce a novel automated approach that combines sub-trees from general taxonomies with specialized seed taxonomies by using specific Natural Language Processing techniques. We provide an extensible and generalizable model for combining taxonomies in the practical context of two very large European research projects. Because the manual combination of taxonomies by domain experts is a highly time consuming task, our model measures the semantic relatedness between concept labels in CBOW or skip-gram Word2vec vector spaces. A preliminary quantitative evaluation of the resulting taxonomies is performed after applying a greedy algorithm with incremental thresholds used for matching and combining topic labels.

Page Composition using PPML as a Link-editing Script

The advantages of a COG (Component Object Graphic) approach to the composition of PDF pages have been set out in a previous paper [1]. However, if pages are to be composed in this way then the individual graphic objects must have known bounding boxes and must be correctly placed on the page in a process that resembles the link editing of a multi-module computer program. Ideally the linker should be able to utilize all declared resource information attached to each COG. We have investigated the use of an XML application called Personalized Print Markup Language (PPML) to control the link editing process for PDF COGs. Our experiments, though successful, have shown up the shortcomings of PPML's resource handling capabilities which are currently active at the document and page levels but which cannot be elegantly applied to individual graphic objects at a sub-page level. Proposals are put forward for modifications to PPML that would make easier any COG-based approach to page composition.

Creating Structured PDF Files Using XML Templates

This paper describes a tool for recombining the logical structure from an XML document with the typeset appearance of the corresponding PDF document. The tool uses the XML representation as a template for the insertion of the logical structure into the existing PDF document, thereby creating a Structured/Tagged PDF. The addition of logical structure adds value to the PDF in three ways: the accessibility is improved (PDF screen readers for visually impaired users perform better), media options are enhanced (the ability to reflow PDF documents, using structure as a guide, makes PDF viable for use on hand-held devices) and the re-usability of the PDF documents benefits greatly from the presence of an XML-like structure tree to guide the process of text retrieval in reading order (e.g. when interfacing to XML applications and databases).

Creating reusable well-structured pdf as a sequence of component object graphic (cog) elements

Portable Document Format (PDF) is a page-oriented, graphically rich format based on PostScript semantics and it is also the format interpreted by the Adobe Acrobat viewers. Although each of the pages in a PDF document is an independent graphic object this property does not necessarily extend to the components (headings, diagrams, paragraphs etc.) within a page. This, in turn, makes the manipulation and extraction of graphic objects on a PDF page into a very difficult and uncertain process. The work described here investigates the advantages of a model wherein PDF pages are created from assemblies of COGs (Component Object Graphics) each with a clearly defined graphic state. The relative positioning of COGs on a PDF page is determined by appropriate "spacer" objects and a traversal of the tree of COGs and spacers determines the rendering order. The enhanced revisability of PDF documents within the COG model is discussed, together with the application of the model in those contexts which require easy revisability coupled with the ability to maintain and amend PDF document structure.

Using SVG as the Rendering Model for Structured and Graphically Complex Web Material

This paper reports some experiments in using SVG (Scalable Vector Graphics), rather than the browser default of (X)HTML/CSS, as a potential Web-based rendering technology, in an attempt to create an approach that integrates the structural and display aspects of a Web document in a single XML-compliant envelope. Although the syntax of SVG is XML based, the semantics of the primitive graphic operations more closely resemble those of page description languages such as PostScript or PDF. The principal usage of SVG, so far, is for inserting complex graphic material into Web pages that are predominantly controlled via (X)HTML and CSS. The conversion of structured and unstructured PDF into SVG is discussed. It is found that unstructured PDF converts into pages of SVG with few problems, but difficulties arise when one attempts to map the structural components of a Tagged PDF into an XML skeleton underlying the corresponding SVG. These difficulties are not fundamentally syntactic; they arise largely because browsers are innately bound to (X)HTML/CSS as their default rendering model. Some suggestions are made for ways in which SVG could be more totally integrated into browser functionality, with the possibility that future browsers might be able to use SVG as their default rendering paradigm.

Vector Graphics: From PostScript and Flash to SVG

The XML-based specification for Scalable Vector Graphics (SVG), sponsored by the World Wide Web consortium, allows for compact and descriptive vector graphics for the Web. SVG s domain of discourse is that of graphic primitives whose optional attributes express line thickness, fill patterns, text size and so on. These primitives have very different properties from those of traditional document components (e.g. sections, paragraphs etc.) that XML is normally called upon to express. This paper describes a set of three tools for creating SVG, either from first principles or via the conversion of existing formats. The ab initio generation of SVG is effected from a server-side CGI script, using a PERL library of drawing functions; later sections highlight the problems of converting Adobe PostScript and Macromedia s Shockwave format (SWF) into SVG.

Mapping and Displaying Structural Transformations between XML and PDF

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. Until recently PDF has been a totally display-based document representation, relying on the underlying PostScript semantics of PDF. Early versions of PDF had no mechanism for retaining any form of abstract document structure but recent releases have now introduced an internal structure tree to create the so called 'Tagged PDF'. This paper describes the development of a plugin for Adobe Acrobat which creates a two-window display. In one window is shown an XML document original and in the other its Tagged PDF counterpart is seen, with an internal structure tree that, in some sense, matches the one seen in XML. If a component is highlighted in either window then the corresponding structured item, with any attendant text, is also highlighted in the other window. Important applications of correctly Tagged PDF include making PDF documents reflow intelligently on small screen devices and enabling them to be read out in correct reading order, via speech synthesiser software, for the visually impaired. By tracing structure transformation from source document to destination one can implement the repair of damaged PDF structure or the adaptation of an existing structure tree to an incrementally updated document.

Encapsulating and Manipulating Component Object Graphics (COGs) using SVG

Scalable Vector Graphics (SVG) has an imaging model similar to that of PostScript and PDF but the XML basis of SVG allows it to participate fully, via namespaces, in generalised XML documents.There is increasing interest in using SVG as a Page Description Language and we examine ways in which SVG document components can be encapsulated in contexts where SVG will be used as a rendering technology for conventional page printing.Our aim is to encapsulate portions of SVG content (SVG COGs) so that the COGs are mutually independent and can be moved around a page, while maintaining invariant graphic properties and with guaranteed freedom from side effects and mutual interference. Parellels are drawn between COG implementation within SVG's tree-based inheritance mechanisms and an earlier COG implementation using PDF.