留言板

Semantics in Liferay

thumbnail
Eirini Voutskoglou,修改在11 年前。

Semantics in Liferay

New Member 帖子: 3 加入日期: 12-11-14 最近的帖子
I'm using liferay 6.1 for the implementation in my master's paper [information systems that is] and although i enjoyed the Vocabulary creation, i really needed an ontology importer. Unfortunately the integration with Ontopia could not work because liferay.ltm is not updated and well.. liferay was lucky that the last version of drupal has not updated the rdf vocabulary importer module for its last version, otherwise i would have quit using liferay..
Liferay should become more semantic friendly and i do not mean through RDFa
thumbnail
Juan Fernández,修改在11 年前。

RE: Need for an ontology importer

Liferay Legend 帖子: 1261 加入日期: 08-10-2 最近的帖子
Hi Eirini:
this is a really interesting topic and I know more community members will love to start a discussion about semantics in Liferay.

What I'd love you to do is to define what is the problem we are trying to solve using semantics in webs built on Liferay and what are the main benefits of implementing this.

After we define those two high level "ideas" we can start thinking in ways to face the problem and step by step break them down into doable tasks and deliverables (plugins for the marketplace? new services in the core? new community product built over Liferay?)

Thanks for your feedback and for taking the time to start this discussion
Juan Fernández

ps: I would rename the thread name to something like "Semantics in Liferay Portal" or something like that...it'll be better to get more interest from the community
Olivier Bonnet-Torrès,修改在11 年前。

RE: Need for an ontology importer

Regular Member 帖子: 162 加入日期: 08-7-11 最近的帖子
Hi Eirini, hi Juan,
Let me jump on the band waggon.
It's been in the air for a while at our facility.
I will be very glad to give a hand on that project.
thumbnail
Juan Fernández,修改在11 年前。

RE: Need for an ontology importer

Liferay Legend 帖子: 1261 加入日期: 08-10-2 最近的帖子
Hi Olivier! It's nice you want to participate!
I am not (at all!) an expert in web semantics, but I'm here to facilitate communication among community members interested in this subject. I'm really interested in what problems you are trying to solve in your company: can you and Erini define this a little bit more so that we can start thinking in general problems to be solved or developed in the product?
Thanks for your feedback!
Juan
thumbnail
Eirini Voutskoglou,修改在11 年前。

RE: Need for an ontology importer

New Member 帖子: 3 加入日期: 12-11-14 最近的帖子
Juan and Olivier thank u so very much for the replies.
I have not finished and presented my paper yet so i can not give many details about what i'm doing but my master was the reason i started studying about semantics.

In the liferay article about taxonomies and folksonomies [which i can not find right now] u mention ontology but this option has nothing to do with an ontology since an ontology must carry individuals and i can not do that through the Categories section.

U r giving an option of Properties for the Categories but i can not really use them for any relationships regarding ontology since i can not have two properties with the same Key field.

The liferay linked module is not updated for latest versions of liferay and it doesn't exactly specifies what it can and can not do.
If it's about the d2rq mapping, well i can do it myself. Do not need a module. If it's about the screenshots portlet of a particular application, well that's a portlet not available.

When i tried integrating liferay with ontopia, it was a fail since liferay.ltm is not updated.
This integration would give me the options of some extremely crucial semantic portlets

All of these were some of the things i wanted and couldn't do [but i still managed somehow to give a semantic air in liferay which i will present in the near future] :-)
thumbnail
Juan Fernández,修改在11 年前。

RE: Need for an ontology importer

Liferay Legend 帖子: 1261 加入日期: 08-10-2 最近的帖子
Hi Eirini:
apparently this topic has been something the community has been talking about previously (see this discussion here)...It's sad, but it's quite common in our community that people discuss about a topic, say they'll contribute something and finally disappear! I hope this conversation leads to some real contributions and product improvements! emoticon

I have contacted the guys that created this integration with Ontopia to see if they plan to update their plugins to the latest version, but the full project seems abandoned. This would be interesting for the community, and the integration seems easy... but someone must update the plugins! I'll update the post if they answer me.

Apart from this I'm pretty sure there are things we can do in the core to improve this but I'll wait for you to present your paper to get the feedback of your experience giving Liferay a "semantic air" emoticon

Thanks for sharing your thoughts with the community!
Juan

Links for further reference:
- Liferay/Ontopia Integration presentation on slideshare
- Liferay/Ontopia Integration howto
- Liferay/Ontopia Integration discussion
- Ontopia semantic portlets for Liferay
Sean Champ,修改在10 年前。

RE: Need for an ontology importer

New Member 发布: 1 加入日期: 12-6-30 最近的帖子
Hello, community forum! I understand that it's been almost a year since this thread was last active, however if it may not be too late to contribute, I think that the topic of using ontologies in Liferay can be focused -- ostensibly -- around SKOS.

Juan Fernández:

What I'd love you to do is to define what is the problem we are trying to solve using semantics in webs built on Liferay and what are the main benefits of implementing this.


If one may provide a brief, albeit perhaps personal response:

Though I've only recently begun to study the internal features of the Liferay architecture -- my having only recently committed to using a single application server as a container for Liferay, namely looking at JBoss AS7, a selection that I think may entail some work in creating UML diagrams describing some of the component frameworks in that architecture, and their applications, for instance ESB, JBI, and JSR 208, before I might be able to get too far dug in with the administrative tasks of running a Liferay portal server on an "Own VPS host" -- I've been reading up about some semantic web technologies, for some period of time -- for instance, reading that knowledge may be regarded as an organizational resource [Alavi2001] if not also or alternately an individual resource, perhaps specifically with regards to studies of science and the humanities. Avoiding some metaphysical arguments with regards to the nature of knowledge, I think there is a debate that may be made in regards to the material nature of knowledge resources, such that perhaps the debate could be of some relevance with regards to web content, in content consumption and content creation flows.

Considering Liferay as a web content creation framework -- on a sidebar, briefly -- I think that although it could seem trivial, but perhaps an analogy can be made in representing Liferay as a sort of content creation crafting table, as in reference to the crafting table motif of the Minecraft game, developed by Mojang. I think that the broader web portal framework, as a model, may seem to represent as sort of evolution of content creation models -- tracing a development timeline from HTTP, then CGI, then to the Servlet Container architecture, then on to the Portlet and Portlet Container framework. I understand that Liferay would have numerous applications in private intranet frameworks. Personally, I've wished to focus more on the public internet as a content presentation venue, of a kind. Personally, in favoring writing as a creative form, I think that the web may be of a quality primarily suited to presentation of literary media. In such a context, I think Liferay can be applied in regards to content aggregation and in regards to creation of original content, though then perhaps not only in production of literary media. In a sense, I think Liferay can be used in developing more decentralized, public social networks, around topics of contextual focus. Perhaps, moreover, a Liferay portal can be integrated with existing social network channels -- in a multi-channel social networking model, with reference to Flipboard -- that the hypothetical, contextually focused web resource might not seem as though it was an island in the broader social web.

In regards to integrating SKOS in the Liferay web content framework, I think that -- ostensibly -- it could be as simple as a matter of extending on the Asset model in Liferay. In regards to matching keywords with existing content, ostensibly, a workflow could be developed in which an ontology representing a model, a controlled vocabulary -- if not a folksonomy or taxonomy of SKOS concepts -- an ontology containing SKOS elements could be uploaded to a Liferay portal, via the portal dashboard. Then, that ontology could be applied onto existing content assets, in the portal -- ostensibly, in a human-verified keyword extraction process, though in regards to large asset databases, perhaps that could be a bit overwhelming, on the verification side. Such a keyword extraction process, it could use an individual keyword extraction tool such as Maui Indexer or a broader framework such as Apache Stanbol. A similar process could be applied in content creation tasks, for suggesting "smart keywords," such as could be applied onto new content items, and those keywords then lastly selected by the individual content creator, before an asset is published via the portal.

Across a portal's full lifecycle, ostensibly, in an application such processes of controlled vocabulary or taxonomy or folksonomy creation and keyword selection, the keyword applications resulting of such a process, it could be of help to portal visitors, for manual content retrieval by portal visitors, and with regards to accurate application of search engine optimization (SEO) techniques -- essentially, in applying a "keyword tagged page" model, as like commonly of the web 2.0.

Juan Fernández:
Hi Eirini:
After we define those two high level "ideas" we can start thinking in ways to face the problem and step by step break them down into doable tasks and deliverables (plugins for the marketplace? new services in the core? new community product built over Liferay?)


Personally -- in all candor -- my having only recently begun to study any details of the Liferay portal architecture, having only recently selected a single portlet server architecture -- I'm not immediately certain of how such techniques could be developed into a single portal application, if not to be developed as to be integrated directly into the web portal itself. In that context, however, tonight I've begun to study the Asset and AssetCategory API in Liferay. I've read about the Liferay Service Builder also -- seen it mentioned, around. I think that I should probably wish to study more, to understand how the assets API is used in its applications within the component services of the Liferay portal, if not to develop a view about asset creation workflow, in regards to Kaleo and other workflow models.

So, perhaps it may be to an initial proposal for how ontologies could be used in Liferay, specifically in focusing about SKOS?

Certainly, GeoSPARQL would be another topic, entirely.
thumbnail
Kevin Emamirad,修改在10 年前。

RE: Need for an ontology importer

New Member 帖子: 12 加入日期: 13-9-4 最近的帖子
Thanks a lot for this contribution Sean, I am a new Liferay developer and I have also some interest on Web Semantics, I didn't knew about SKOS. So I will read more around this ;)