Software that understands content makes a CMS significantly more valuable. Due to automated reasoning, no more manual work will be needed, for example by content authors to display related content. Website visitors will benefit from context-aware search that yields the best results from vast amounts of content. By combining Magnolia CMS with salsaDev’s technology, users save time and maximize the value of content.

What you will learn:

• How to quickly install and configure the searchBox solution for Magnolia CMS
• How to customize the searchBox integration and ensure high performance
• Typical scenarios and the business benefits of semantic search technology

This Webinar is extremely useful for CMS developers of all levels, as well as decision makers interested in deploying a semantic CMS.

July 5/6 – Paris. Great turn-out earlier this summer, for the IKS Paris Workshop, as more than 90 industry experts participated in heated discussions on how semantic technologies contribute to content management experiences. As a salsaDev (semantic vendor and IKS early-adopter) crew member, I had the pleasure to present our perspective, highlight how we’re leveraging semantic technologies including IKS Apache Stanbol and our own, and discuss business with peers.

In this wrap-up post about the salsaDev experiences, I’d like to highlight some specific customer-centric use-cases where a semantic CMS really makes sense for organizations, and which features are most useful in this process. During the workshop, some opinion leaders and innovators including Stéphane Coisier (Jahia) and Janus Boye (Janus Boye) clearly highlighted the “lack of maturity of semantic technologies” together with the “missing semantic applications and the high expectations of customers”. While I agree with both statements, perhaps the following elements are also missing:

1. some customers may think of semantic technologies as “black magic” and hence anything is possible with it (thus making it difficult to identify a core set of mainstream functionalities, c.f. IKS User-Stories), and
2. sometimes the less the technology is visible, the better is the application. In short, there might be a shortage of “killer semantic applications”, and probably the IKS Semantic UI/X competition, planned for November this year is the right step towards digging-out and developing such a killer semantic-out-of-the-box application!

In the salsaDev IKS Early-Adopter proposal, we proposed to take this out of an academic / R&D environment, and deploy the combined integration for one specific customer and a few use-cases. To note, as is explained in the EA proposal, salsaDev is querying several engines, incl. Apache Stanbol and it’s own salsaAPI v.2.1, and if applicable will merge results from engines based on advanced heuristics. The features explained below detail those particular-ones where the Apache Stanbol engine is being queried (some screenshots are also made available to highlight the front-end impact for the users).

Similarity-based document search (S09 IKS User-Story) This Apache Stanbol feature is extremely useful when i) navigating through individual search results, as it provides a list of contextually related documents and articles, and ii) creating a new content, in order to check which documents are already available in the system (avoiding content duplication).

Automated document categorization (S10 IKS User-Story) This Apache Stanbol feature represents the possibility to add a categorization framework out-of-the-box. As such, the customers may leverage existing categorization schemas, or define their own set of n-level categories. Then, the system will be able to automatically categorize all existing and incoming content directly, and continuously learn and improve its categorization based on user-feedbacks.

Semantic tagging (coming soon – S05 IKS User-Story) Although this Story is not yet implemented by salsaDev, it will soon as many customers are requesting it. This story makes it possible for users to offer “a list of tags that describe my content, and a list of links to entities that the text talks about”.

One of the key points of salsaAPI is how easy it is. Let’s see for example how you can index some relevant content and then automatically match it with other items sharing a similar meaning, whatever their length is and even if there is no prior cross-references or annotations between all these objects.

Please note that salsaAPI is based on an HTTP/REST model using XML for communication, so you need some basic knowledge of such technologies before being able to get started.

1. Build appropriate XML files with your documents (e.g. PDF, Word, raw text) and index them with HTTP/REST method POST /document. Your documents are then indexed in salsaAPI in your own repository and are identified with your own unique identifiers (e.g. the primary key from your database, the file path from your file system, the URL of your web article, etc.).
2. Find content related to one of your document with the HTTP/REST method GET /related/{uid}, where {uid} is the unique identifier of your document.
3. The response is a XML file which itself contains unique identifiers pointing to similar data.

There are plenty of use cases you can leverage, including automatically providing shortcuts to your old archived content, to other content posted by other users, or to automatically match your current content item with the one of third party knowledge bases.

Leveraging other features provided by salsaAPI’s semantic engine is just as easy (e.g. search, keyword extraction, or auto-categorization). So why not give a try: the salsaAPI free trial is waiting for your calls and a more detailed documentation is available in our wiki!

The salsaDev NipTech 079 Interview is available on the NipTech website! It was a great interview with Benoit Curdy, Podcaster of Niptech.com, talking about the various semantic activities of salsaDev. All the podcasts of NipTech are made available on the iTunes store. Most (if not all) of the interviews are conducted in French.

In short, Benoit was interested to learn more about our professional activities, together with the challenges, advantages and disadvantages of building a successful start-up company in Western Switzerland. We discussed at length all the various Swiss institutions promoting Research & Innovation in Switzerland, including AlpICT, VentureKick, VentureLab and many others. Stay tuned for more podcasts!

Trend-Setting Product awards began in 2003. This year, more than 800 products were assessed by a judging panel, which consists of editorial colleagues, analysts, system integrators, vendors, line-of-business managers and users. All products selected demonstrate clearly identifiable technology breakthroughs that serve the vendors’ full spectrum of constituencies, especially their customers.

The salsaAPI 2.1 provides a unique and comprehensive set of semantic services, including semantic indexing, concept search, categorization and finding related documents/articles. The salsaAPI offers a powerful RESTful API and can be used online or onsite, and can be easily embedded within any 3rd party sites or applications. For instance, plugins are available to deliver an advanced search experience into leading Content Management Systems & Corporate Portal Servers.

Available modules include

 Liferay Portal EE: www.liferay.com

 Magnolia-CMS: www.magnolia-cms.com

The Liferay West Coast Symposium 2011 (#LRWCS2011) is coming to an end in Anaheim, California. It’s been an exciting 2 days, kicked-off by the back-to-back announcements by Liferay on its 6.1 release and launch of the Liferay Marketplace. The objective of the marketplace will be to enhance Liferay’s lightweight platform with enterprise applications and plugins that augment Liferay’s core functionalities.

salsaDev is really excited to be among the contributing ISVs to the Marketplace platform alongside Alfresco, Roambi, SocialEkwity and others. The salsaDev semantic enterprise applications will enable organizations to index all their content within the portal, and benefit from innovative semantic features including: Auto-categorization, Full-text & concept search, and suggested/related content.  The salsaDev tech team will provide support & training to the Liferay developer community wishing to beta-test and leverage the semantic apps during the launch phase of the Marketplace and afterwards.

The Liferay 6.1 release introduces a number of new features to expand the WCM facet of the product, document management capabilities and support for data lists and workflow forms. A faceted search API is also made available courtesy of SavoirFaireLinux, an open-source consultancy firm based in Montreal, Canada.

The Liferay WCS2011 has been extensively covered by CMS journalists including @cmsreport, @cmscritic and @CMSWire. salsaDev is a Liferay WCS2011 Marketplace exhibitor.

salsaDev presented its latest product, salsaAPI v.2.0. together with the inntegration into the Apache Stanbol/FISE semantic stack IKS Amsterdam workshop. As an IKS early-adopter, salsaDev is at the fore-front of the integration of semantic technologies including Stanbol and the salsaAPI into Content Management Systems.

The ultimate objective is to “semantize your CMS”, making it possible to leverage semantic technologies and their associated features directly into CMS stacks. In as a transparent and efficient way as possible for the end-users. Some preliminary results show promising results, with salsaDev’s integration into Jahia and Liferay. Zaizi also demonstrated it’s OpenCalais integration directly within Alfresco.

More to come in IKS next workshop’s with customer case-studies and “killer semantic applications”!

Interesting article in AlpICT Bilan on Cloud-based swiss start-ups: “Their names are Paper.li, sobees, nViso, salsaDev, Fontself, or HyperWeek and Camptocamp. They are the rising stars of the Web 2.0 in Western Switzerland. Less than three years old, they are already experiencing growth rates in their Internet traffic of three digits or more.”

“At salsaDev, a Geneva-based company that develops search engines using cutting-edge semantic technologies, CEO Nicolas Gamard adds: “The big advantage of cloud computing is that what was previously a large capital investment turns into a smaller operational expenditure. As such, it’s easier for Internet start-ups to convince venture capitalists to invest money in core business activities rather than in servers and data centers.”

This blog post describes the transition of technologies from academia to industry, specifically in the semantic search realm. This first post discusses what and why the differences between academia and industry exist, while the second post will more fully develop the pros and cons of various semantic technologies, in and out of academia.

To sum the difference between the two paths, it comes down to a single word: utility. The goal of most enterprises is monetize the solution of problems which people face, as a result creating free-time and/or reducing the amount of effort to complete a goal.

In the case of search solutions,  the International Data Corporation (IDC) has recently conducted studies which clearly indicate the problem of information access:

1. Knowledge workers spend from 15% to 35% of their time searching for information.

2. Only 21% of respondents said they found the information they needed 85% to100% of the time

3. 40% of corporate users reported that they can not find the information they need to do their jobs on their intranets.

4. 90% of the time that knowledge workers spend in creating new reports or other products is spent in recreating information that already exists.

providing a great motivation for improved search capabilities, and even a large space for a significant paradigm shift of how users access their information.

While there already exist a large number of papers published in the information retrieval realm, it is safe to say that not all of them take into account the challenges and needs faced by industry implementation.

The main points to be considered by “industry friendly” algorithm are:

1. Scalability: How well can the algorithm handle large datasets, wow well does it support the addition of new documents, and can it work on industry standard hardware configurations (i.e. memory usage)?

2. Efficiency: Is the response time acceptable for all use-cases, especially the most common (i.e. search should be fastest, while addition of a new document can be a bit slower)? Can it run for months without painful maintenance or upkeep?

3. Robustness to user error:If inproperly used, what are the negative side effects? Traditionally this is most concerning in supervised systems, which ask for a training set and then require hours of computation. What if a layman were to train the algorithm, would it still operate close to expectations?

It is not too difficult to imagine a fresh cutting edge algorithm to violate some of the above requirements. In many of the algorithmic competitions, we see computation time reaching into days and weeks in an attempt to squeeze out a few more percentage points, but any change (such as document addition) to the data requires the algorithm to start from scratch. We see carefully created training sets, created by experts, used to boot strap supervised algorithms, which would be expensive in both time and money to create on a per client basis.

While many ideas are proposed which are intellectually stimulating and provocative, only the algorithms which work, and work well, make it into an industry setting.