Semantic Links Wiki

Now the wiki site is created to accumulate the information about the Semantic Links and about the Semantic Web. You can find it here: http://www.semanticlinks.net/wiki.

One year in Live

The Semantic Links project is one year in live. Now It supports the FireFox. Thanks to Firebugs! Also I implemented a possibility to inject the Semantic Links functionality to any site. I'm still working on the Class Designer and new widgetes. Let me know if you'd like to attend the project!

Furebug - great web development tool

This is an excellent add-on for the FireFox. The Furebug lets debug front-end stuff with pleasure because it provides such cool features as:

• JavaScript debugger
• Nice watch panel that shows all about JavaScript objects, functions, variables
• DOM Viewer
• CSS Viewer
• Dynamic HTML Viewer that can higliht elements on the FF when your mouse is over HTML tag
• Inspector - to select any object on the rendered page and show it's HTML
• Console
• Editor
• Features to speed up your code

Very useful tutorial is here. Also you can find some videos about Firebug.

To Open or not to Open?

Well, I'm thinking about make the Semantic Web project an open source. It seems that it can push the project.

Yahoo! Proposed they mashup builder. The main feature that you can design your mashup with graphical designer and use some predefined operators and modules.

The mashup is a pipe in Yahoo! terms. So you can create your pipe with use:

• Inputs from Yahoo Search, Fetch, Google Base, Flickr.
• User inputs: numbers, strings, URL, location, …
• Operators – they help to process data: create loops, counters, filters and other.
• Other modules and operators.

Very interesting is that you can reuse other pipe in your own pipe.

Mostly pipes are suitable to analyze RSS feeds. This is because Yahoo provides several cool modules that work with feed. For example I created a pipe that do the following:

1. Gets user input (search text)
2. fetch feeds from:
1. http://del.icio.us/rss/tag/semantic
2. http://del.icio.us/rss/tag/semanticweb
3. http://del.icio.us/rss/tag/rdf
3. search for “search text” in every entry of feeds

The picture with this simple pipe is shown below.

Pipes can be useful to search what you need across feeds with similar semantic. Especially if you search every day.

It’s not so tricky to create your own pipe. Also I’m sure that number of allowed modules will grow and more complex pipes will be possible just as 1-2-3.

Start from example pipes to understand how they work http://pipes.yahoo.com/tags/example/

On January 16, Geni launched. It allows people to create quick and beautiful family trees.

That’s really simple but valuable idea. A kind of social network built on the sibling connection. Just imagine that next step of the Web 2.0 is to interconnect different social networks. In this case we’ll have “real digital persons” with a lot of connections of different types. This will be a many-dimensional web.

Demo of the new WEB ;)

Well, now I pushed portion of Semantic Links directly to my blog (see SemanticLinks Demo section). What’s this? This is a catalog of innovations directly on my blog. And lots of other ideas can be implemented in the same manner ;). I’d like to tell how it works. Earlier I decided to create the catalog of innovations based on the Semantic Links engine. So I created several classes for this (I used Protege, but now I’m going to create own class designer). These classes are:

• Catalog (the catalog itself). It contains only list of Topics and HeadTopics.
• HeadTopic – is a simple class that visualizes a title and some html page as a description.
• Topic – This lists innovation companies within some area (IT, Astronomy, etc…).
• Provider (lets say innovation company). It contains all information about innovation company including:
• List of projects with they descriptions (showcase)
• Attributes of the company (address, phone, etc…)
• Tags
• Project – Just description of project.

Classes are depicted below:

After that I tuned look & feel for these classes. I just specified what widgets to use for attributes. For TopicList I specified MasterDetailWidget and specified Topics Attribute as a master list. ListWidget will be used to visualize and provide behaviors for Topics (and HeadTopics) list. The detail area stays empty. This will be loaded when some item selected in a master list.

I don’t want to describe all classes, but I’ll describe most complex class in my ontology: Provider. Provider contains 10 attributes and I specified widgets for them:

• ImageWidget for ImageLabel attribute
• LabelWidget for:
• Title
• Country
• City
• Email
• SiteURL
• TextWidget for Description
• ListWidget for:
• Phones
• Showcase
• Tags

Most of them are depicted below:

That and now I have a result that can be simply checked if you have IE browser. Just press button in the “SemanticLinks Demo” section (Right Top area of the blog). If you have Microsoft IE browser then you can play with it and see the results.

Blog about Semantic Web and Web 2.0 in Russian

Блог сайта посвящен дальнейшему этапу развития веба - семантическому (умному). В блоге рассматриваются такие революционные направления как семантические поисковые системы, ajax-приложения, виджеты, гаджеты - являющихся переходящим звеном с Web 2.0 на новую концепцию Web 3.0. SemanticFind

Some cool JavaScript controls

Well, Now I'm working on the class designer for my ontology based service. While working on it I found some interesting JavaScript controls. Take a look:

What the Semantic Links similar to

Hello, colleagues! Thank you for several comments. I'll try to make the goal of the project more clear. At the moment two parts of project exist: 1. The engine and the front-end widgetes that make it possible to visualize ontologies and process them 2. Currently developed ontology that realizes a catalog of innovations. Of course there a number of similar services: - google base - linkedin and of course realizations of FOAF ontology. Regarding linkedin and FOAF realizations. I'm using LinkedIn but as I see it can only process fixed ontology (that seems to be similar to FOAF). My goal is make an open environment for different custom ontologies and give a tools to create and process these ontologies. So I would prefer to compare the Semantic Links with the Google base. As I see Google base doesn't use any Semantic Web technique (http://www.informit.com/guides/content.asp?g=xml&seqNum=280&rl=1) I trying to build the same thing on OWL.

About the Semantic Web Services

Now I'm exploring the web to extract interesting stuff for SWS. To read about SWS and to find tools for development you can see http://www.daml.ri.cmu.edu/tools/index.html. Also it's an interesting plugin for Axis: http://www.mcj.de/mirrors/owlsplugin/. And useful link from mindswap: http://www.mindswap.org/2004/owl-s/

Catalog of innovations

Dear colleagues, I decided to go with more precise idea. For the beginning I want to start the catalog of innovations. Let proof the innovation Semantic Web and Web 2.0 technologies on the task connected to innovations ;) Well, now I invite you to put your cards (budges) to the catalog. Currently you can:

• add your card
• describe your card with tags
• add existing tags
• define your tags
• add your logo, short description, contact info
• add list of your ideas, projects or products. Every such description can contain:
• picture
• short description
• manage your cards

After you’ll add your card just write me a publishing request to support@semanticlinks.net. This is necessary to prevent spam in the catalog ;) I’ll be very thankful for your comments and ideas. Please feel free to comment blog posts and write to a forum (www.semanticlinks.net/forum).

How to become published in catalog

Dear colleagues! To make your information globbaly accessible you may: 1. Write me and I'll create your card and publish it for a little fee (you'll have a possibility to manage your information). 2. You may register and create your card by yourself and write me a request to publish your card (for a little fee as well ;) ). Regarding money: this fee is necessary to continue the project. In nearest future I'll implement publishing service. So you'll have a possibility to publish your cards (for a little fee as well ;) ) Thanks a lot! P.S. Next time I'll write how you can use the functionality of Semantic links on your sites.

Today my hosting was down for several hours :( please don't kill me!

The major ideas of the Semantic Links area

There are three main points in the Semantic Links area: 1. Web Ontology Language (OWL) is used to represent structure of resources; 2. The resources should be easily readable and accessible for users; 3. The resources should be understandable by services. So I use OWL to represent data. You can see the example here but please don’t subtilize ;) Such data is really understandable by the software agents or services but not for people. For people it’s more appropriate to have something like picture 1 to understand the hierarchy on nouns:

And something like picture 2 to understand the structure of resources and to manage resources:

These screenshots show how the Protégé represents OWL classes and instances. To represent the instance Protégé uses widgets conception. For every slot in OWL class it will define default user interface widget of corresponded type. User can change the default widget for something else. E.g. if you have some text slot, it can contain the URI of image. So the user can use an image widget to view image by URI. The Semantic Links area utilizes the same idea. So I have a main OWL model that contains the meaning information (catalog.owl) and the model of UI representation (uidata.owl). The model of UI representation contains definitions of UI elements. So I used OWL to represent the meaning data and UI definitions. The UI definitions determine the representation of resources. The following OWL example expresses such approach:

<Widget rdf:ID="uiformdescriptions_RDFResource_38"> <Name rdf:datatype="http://www.w3.org/2001/XMLSchema#string">http://reflections/bcatalog#Service</Name> … <Widget rdf:ID="uiformdescriptions_RDFResource_22"> <Left rdf:datatype="http://www.w3.org/2001/XMLSchema#int" >20</Left> <Name rdf:datatype="http://www.w3.org/2001/XMLSchema#string" >http://reflections/bcatalog#ImageURI</Name> <Top rdf:datatype="http://www.w3.org/2001/XMLSchema#int">80</Top> <WidgetRealizationName rdf:datatype="http://www.w3.org/2001/XMLSchema#string" >IconWidget</WidgetRealizationName> <Caption rdf:datatype="http://www.w3.org/2001/XMLSchema#string" >Image</Caption> <Width rdf:datatype="http://www.w3.org/2001/XMLSchema#int" >300</Width> <Properties rdf:resource="#uiformdescriptions_RDFResource_15"/> <Height rdf:datatype="http://www.w3.org/2001/XMLSchema#int" >60</Height> </Widget> … </Widget>

This example shows that widget for http://reflections/bcatalog#Service class contains widget that represents http://reflections/bcatalog#ImageURI slot. To represent slot IconWidget will be used. In the following picture the UI of http://reflections/bcatalog#Service depicted.

The forerunners of the project

The idea of the project was born after a lot of work for metadata based development. Currently there are a lot of technologies and platforms that provides different Domain Specific Notations to develop business software. Now I’m involved into one such project “Murometz” (www.realbh.ru on Russian only) which is based on the following ideas: 1. Metadata (ontology) is used to describe Domain Entities and they specific relations 2. Generators are used to generate the VB.Net code for: Domain Entities (object model) Representation of Domain Entities (user interface model) 3. Number of services (together they build up a platform “Murometz”) are used to process Domain Entities Security Persistence Search Replication Etc… In the Semantic Web world there are lots of tools for ontology development. And there is obvious that ontologies help program services and agents to understand the sense of information and relations between different resources. Now it’s really important to find a way how bring together ontologies, information resources and services and deliver they common power to end-user and business. First of all it’s necessary to provide every user with possibility to publish and manage his information in an ontology based manner. When amount of such information will get to a critical mass the semantic services and agents will grow extremely. And we’ll have a possibility to take advantages of such services. So this is the major idea.

How this can be done now? Well there is a number of tools that provide user with possibility to create own ontology and to create information resources in terms of such ontologies. I like the Protégé (http://protege.stanford.edu/) tool which is really powerful. But the Protégé is a java program oriented for developers. You should download it before you can add create some resource. Moreover you should deploy your resource to some open area. There is a WEB based project that provides some of Protégé functions on the web (see pOWL project http://powl.sourceforge.net/). The Semantic Links project will provide similar functions but it should be simpler for end-users. Also it should provide powerful functionality oriented to Web developers, service developers and ontology developers.

Diving into the Semantic Web

Dear colleagues! This post starts my journal about the Semantic Links project (www.semanticlinks.net). In my journal I plan to tell about the project itself, about the Semantic Web technologies and about some development know-how connected to the Semantic Web techiques. For me it's really obvious that the Internet is going to be a great brain with lots of semantically linked resources and smart services (or agents) that can explore such resources. So I started the Semantic Links project to introduce an open area for placing and linking any structured information resources. What is Semantic Links for? Well, I’ll illustrate it with one example: The catalog can contain information about flights, transfer services and hotels. Planning your trip you can consult the planning service. This service will explore available information basing on the semantic links and compare it with your preferences. After that the service will advise you what flight to prefer, what transfer services and what hotels will be best for you. If those companies have Internet ordering services than you have a possibility to order everything in a moment. The main point here is that everybody should have a possibility to create his own information resources and introduce these resources into common semantic network for further semantic processing. I believe that the number of semantic services in the Internet will grow. So these services can utilize the Semantic Links to solve they tasks.