Paux Wiki

PAUX is a software to develop, manage and individualize publications of dynamic content by linking reusable semantic content objects semantically.

By this innovative method, PAUX takes a new path of knowledge representation by linking single objects such as words, sentences, pictures, persons etc. to a complex but dynamic network. Therefore, PAUX can be classified partially in the area of existing systems like CMS/ECMS, Wiki, LMS or LCMS. The semantic content objects of PAUX represent knowledge in its full complexity. This makes knowledge available as filterable content for 

    * Websites,
    * semantic Wiki,
    * detailed-evaluated eLearning and
    * individualized print media.

PAUX is a knowledge management system written in Java and addresses enterprises (also publishing houses) as well as colleges, offices and other authorities. It is free of charge in the context of a development partnership.

History and distribution

PAUX was founded in the year 2000 by Michael Dreusicke and is since then under constant development and improvement. In 2001 a beta version was written in C++. In 2005 PAUX was completely re-developed based on the database-framework MyTISM with full support of Java. The success of PAUX lies in its high-granular data model, which stores and connects words, sentences and paragraphs based on data base objects. Development partnerships exist with the financial management Rheinland Pfalz in Germany (since 2005) and with the medical faculty of the Technical University of Munich (since the completion of the last version in 2006).

Differentiation to other content management systems

PAUX does not store HTML pages, but content objects themselves, so-called PAUX-Objects (PAUX-Objects can be words, sentences, headings, pictures, persons etc.). 

    * PAUX-Objects are not linked by hyperlinks, but by “PAUX-Links”.
    * PAUX-Objects are called “semantically” due to their linkage with other PAUX-Objects. Those links represent the full context of the PAUX-Object, and they allow the semantic search and filtering for other PAUX-Objects. These objects can be filtered on the basis of it's own attributes as well as attributes of the linked objects.
    * PAUX-Links are called semantic, due to the fact that they can be rated and extended with attributes (see below). Based on this network, PAUX-Objects can be searched and filtered by their semantic context (e.g. user-group-specific or context-oriented).

PAUX-Objects

PAUX-Objects represent the content itself and are stored as discrete data base objects. Due to the high granularity, the content is completely reusable (Reusable Learning Objects).

Container objects (are able to contain other PAUX-Objects) 

    * modules, headings, categories of headings, lessons, tests, sentences
    * ratings (for each PAUX-Object there can be a set of weighting objects)
    * faculties, subjects and subareas
    * sessions, runs
Example “info-object”: Info is an autonomously formulated declarative sentence, which is supplemented with meta data (11 further text fields) and semantically linked with 44 other classes. Example “heading-object”: A heading contains several info-objects, it has 21 further text fields and is semantically linked with 37 other classes.

Standard objects 

    * information parts (words, intermediate questions, intermediate answers, multiple choice answers, cloze answers)
    * Immersions
          o Footnotes (essays, judgements, case studies, data media, books (textbooks and comments)
          o Media (documents, pictures, animations, audios, videos)
          o persons (person can have specific roles)
    * natural persons
    * Legal entities of the public law (bodies, institutes, foundations) and private law (bodies, foundations)
          o events
          o products
          o hyperlinks
    * weighting objects (target groups, difficulty level, relevance, quality, reasons, notes)

PAUX-Links

The PAUX-Objects are not linked by hyperlinks, but by independent semantic data base objects, so-called PAUX-Links. Those can be searched and filtered semantically for other PAUX-Objects e.g. target-group-specific or context-oriented. The linkage based on PAUX-Links is called semantically, due to the fact that they can be weighted according to different weighting criteria ("weighting objects", see above.).

The linkage of the PAUX-Objects through PAUX-Links causes that 

    * contents becomes semantically searchable,
    * linkages can be listed automatically,
    * the semantics of the objects remain preserved,
    * the semantics of the linkages remain preserved,
    * back relations can be listed automatically.
Example: Info- and heading-objects are connect via a n:m-relationship (headings can contain a set of different info-objects and one info-object can occur in a set of different headings) and are linked by the PAUX-Object “InformationWithHeading?”. It contains the ID of one info-object, the ID of the heading-object and an arbitrary number of weighting-objects (“InfoWithHeadingWeighting?”), which says, for which target group this information is how difficult in this heading is, as well as the level of relevance and reason for this weighting and notes. So the functionality of a PAUX-Link goes far beyond the one of a usual hyperlink.

Differentiation to other Wiki Engines 

    * PAUX is a Semantic Wiki, which is not based on static HTML pages, but instead consists of high dynamic documents, which are provided at runtime from semantic micro objects (“PAUX-Objects”), which are semantically linked by “PAUX-Links”. Therefore content can be filtered extensively (e.g. by target group).
    * One word, therefore, can not only be linked to one (1) target (like it is done via hyperlinks or in mindmaps), but it can have any desired number of objects semantically connected to it (Objects are sorted according to their class, and back relations can be provided automatically). PAUX forms thereby a multidimensional mindmap (and/or due to the semantics of the linkage more exactly: a concept map).
    * The user does not leave the source object by pursuing a PAUX-Link, but can explore via Mouse Over a whole set of objects linked with the word by so-called PAUX-Menus based on Popups (instead of jumping to a single page with heterogeneous information).

Differentiation to other eLearning platforms 

    * PAUX sentences (statements) can also be formulated as a question and answer (cp. index cards including interposed questions and interposed answers), multiple choice and clozes. Furthermore they can be connected to a set of deepenings (footnotes and media types, see above “PAUX-Objects”).
    * The question and answer alternatives have in each case a response time, points (bonus and malus), reasons (why the anser is true or false) and target-group-specific deepenings. With this method, the user is successively led from the passive knowledge to the gain of active knowledge.
    * The user behaviour is evaluated in detail:. Examiners can check for each user, when he/she has examined which sentence, and has the option to monitor the individual learning level and success rate. Even the selection of the answer alternatives and whether the user has investigated deepenings offers will be logged. The sentences with the most correct or wrong answers can be displayed. Thus examiners see, where the users have their most difficulties, and where further support is needed.

Differentiation to other print-authoring tools 

    * PAUX contents can be filtered and searched semantically(thus according to content based criteria, based on PAUX-Objects and PAUX-Links), so they can be easily and fast prepared didactically and linked.
    * Contents can be filtered semantically and printed separately (e.g. for selected topics only the multiple choice [test]]s for a certain target group).
    * The format of the PAUX-Objects and page layouts can be controlled in detail over css stylesheets.
    * Database reporting tools (like e.g. iReport) can automatically create articles, magazines or books for the pre-press.
    * Provided output formats are : PDF, HTML, XLS, CSV, LaTeX and XML.

Function range 

    * 3-Tier-Database-Framework - PAUX can be easily enhanced, because it’s java programming code is generated by the database framework MyTISM.
    * High-capacitiy database - the data is stored into the open source data base PostgreSQL. The data can be edited online and offline (synchronizing servers).
    * Integration in existing systems - With Java and PostgreSQL PAUX uses exclusivly established standard components and can be integrated therefore optimally into existing systems.
    * Object orientation - PAUX-Objects and PAUX-Links are fully object-oriented so that properties are automatically inherited by parent classes to their child classes.
    * Import function - Via text import texts can be divided automatically into their components (paragraphs, sentences, words) and linked according to the output text. If the text contains page ]]number]]s, footnotes are created automatically to each heading and to each sentence of the text. Pictures can be imported individually or via whole directories.
    * Right administration - Authors can read (according to standard) contents of all authors and create own linkages. An author can change or delete however only objects of members of his assigned author-group. The right administration goes down to the attribute level and scripts can be applied. Therefore, differentiated individual authorizations can be created and confidential contents can be protected.
    * History – Modifications of all objects will be logged in detail, so that the authorship of all contents is completely verifiable.
    * Deletion protection – Even if objects will be deleted by an author, the object will remain in the database and can be restored on demand.
    * Notification system - Authors and users can be notified, when a container-, a single or a linked object has been changed.
    * Templates - The appearance of the PAUX-Objects and Wikis is controlled via css templates.
    * Internationalization - PAUX is multilingual. After log-on the user can select his preferred language.

EditText of this page (last edited October 9, 2006) or FindPage with title or text search