Curated for content, computing, and digital experience professionals

Year: 2012 (Page 6 of 7)

Speaker proposal update

Thanks all for the speaker proposals!

Next step is a preliminary organization by the program committee to see if we have all the topics covered.

If you have submitted a proposal remember that it may be a few weeks before a decision is made, but we will keep you posted here on our overall progress.

W3C Launches Linked Data Platform Working Group

W3C launched the new Linked Data Platform (LDP) Working Group to promote the use of linked data on the Web. Per its charter, the group will explain how to use a core set of services and technologies to build powerful applications capable of integrating public data, secured enterprise data, and personal data. The platform will be based on proven Web technologies including HTTP for transport, and RDF and other Semantic Web standards for data integration and reuse. The group will produce supporting materials, such as a description of uses cases, a list of requirements, and a test suite and/or validation tools to help ensure interoperability and correct implementation.

A rarity these days – an announcement that used ‘data’ instead of ‘big data’! And the co-chairs are even from IBM and EMC.

Search Engines; They’ve Been Around Longer Than You Think

It dates me, as well as search technology, to acknowledge that an article in Information Week by Ken North containing Medlars and Twitter in the title would be meaningful. Discussing search requires context, especially when trying to convince IT folks that special expertise is required to do search really well in the enterprise, and it is not something acquired in computer science courses.

Evolution of search systems from the print indexes of the early 1900s such as Index Medicus (National Library of Medicine’s index to medical literature) and Chemical Abstracts to the advent of the online Medical Literature Analysis and Retrieval System (Medlars) in the 1960s was slow. However, the phases of search technology evolution since the launch of Medlars has hardly been warp speed. This article is highly recommended because it gives historical context to automated search while defining application and technology changes over the past 50 years. The comparison between Medlars and Twitter, as search platforms is fascinating, something that would never have occurred to me to explore.

A key point of the article is the difference between a system of search designed for archival content with deeply hierarchical categorization for a specialized corpus versus a system of highly transient, terse and topically generalized content. Last month I commented on the need to have search present in your normal work applications and this article underscores an enormous range of purpose for search. Information of a short temporal nature and scholarly research each have a place in the enterprise but it would be a stretch to think of searching for both types via a single search interface. Wanting to know what a colleague is observing or learning at a conference is very different than researching the effects of a uranium exposure on the human anatomy.

What have not changed much in the world of applied search technology are the reasons we need to find information and how it becomes accessible. The type of search done in Twitter or on LinkedIn today is for information that we used to pick up from a colleague (in person or on the phone) or in industry daily or weekly news publications. That’s how we found the name of an expert, learned the latest technologies being rolled out at a conference or got breaking news on a new space material being tested. What has changed is the method of retrieval but not by a lot, and the relative efficiency may not be that great. Today, we depend on a lot of pre-processing of information by our friends and professional colleagues to park information where we can pick it up on the spur of the moment – easy for us but someone still spends the time to put it out there where we can grab it.

On the other end of the spectrum is that rich research content that still needs to be codified and revealed to search engines with appropriate terminology so we can pursue in-depth searching to get precisely relevant and comprehensive results. Technology tools are much better at assisting us with content enhancement to get us the right and complete results, but humans still write the rules of indexing and curate the vocabularies needed for classification.

Fifty years is a long time and we are still trying to improve enterprise search. It only takes more human work to make it work better.

One week till Gilbane Boston speaking proposals deadline!

Every year we get a last minute rush of speaking proposals for Gilbane Boston, and then… we get tons of emails asking when the deadline is, and then… we get requests for an extra day or two, and then… well, you get the picture. You’ve got a week, but why wait till the weekend!?

The deadline this year is May 14th. Here are the relevant links:

Embedded Search in the Enterprise

We need to make a distinction between “search in the enterprise” and “enterprise-wide search.” The former is any search that exists persistently in view as we go about our primary work activities. The latter commonly assumes aggregation of all enterprise content via a single platform OR enterprise content to which everyone in the organization will have access. So many attempts at enterprise-wide search are reported to be compromised or frustrated before achieving successful outcomes that it is time to pay attention to point-of-need solutions. This is search that will smoothly satisfy routine retrieval requirements as we work.

Most of us work in a small number of applications all day. A writer will be wedded to a content creation application plus research sources both on the web and internal to the enterprise in which writing is being done. Finding information to support writing whether it is a press release, marketing brochure or technical documentation to accompany a technical product requires access to appropriate content for the writer to deliver to an audience. The audience may be a business analyst, customer’s buyer or product user with advanced technical expertise. During any one work assignment, the writer will usually be focused on one audience and will only need a limited view of content specific to that task.

When a search takes us on a merry chase through multiple resource repositories or in a single repository with heaps of irrelevant content and no good results, we are being forced into a mental traffic nightmare, not of our own making. As this blog post by Tony Schwartz reminds us, we need time to focus and concentrate. It enables us to work smarter and more calmly; for employers seeking to support workers with the best tools, search that works well at the point of doing an assignment is the ultimate perk. I know how frantic and fractionated my mental state becomes as I follow one fruitless web of links after another that I believe will lead me to the piece of information I need. Truthfully, I often become so absorbed in the search and ancillary information I “discover” along the way that sight of the target becomes secondary.

New wisdom from a host of analysts and writers suggests that embedded search is more than a trend, as is search with a specific focus or purposeful business goal. The fact that FAST is now embedded with and for SharePoint and its use is growing principally in that arena illustrates the trend. But readers should also consider a large array of newer search solutions that are strong on semantic features, APIs, integration options, and connectors to a huge variety of content that exists in other application repositories. This article by James Martin in CIO, How to Evaluate Enterprise Search has helpful comments from Leslie Owens of Forrester Research and the rise of connectors is highlighted by Alan Pelz-Sharpe in this post.

Right now two rather new search engines are on my radar screen because of their timely entrance to the marketplace. One is Q-Sensei, which has just released their version 2.0. It is an ontology-based solution very much focused on efficiently processing big data, quick deployment, and integration with content applications. The second is Cambridge Semantics with its Anzo semantic solutions for analyzing and retrieving business data. Finally, I am very excited that ISYS was the object of an acquisition by Lexmark. It was an unexpected move but they deserved to be recognized for having solid connector/filter technology and a large, satisfied customer base. It will be interesting to see how a hardware vendor, noted for print technology, will integrate ISYS search software into its product offerings. Information retrieval belongs where work is being done.

These are just three vendors poised to change the expectations of searchers by fulfilling search needs, embedded or integrated efficiently in select business application areas. Martin White’s most recent enumeration of search vendors puts the list at about 70; they are primarily vendors with standalone search products, products that support standalone search or search engines that complement other content applications. You will see many viable options there that are unfamiliar but be sure to dig down to understand where each might fill a unique need in your enterprise.

When seeking solutions for search problems you need to really understand the purpose before seeking candidate vendors. Then focus on products that have the same clarity of applicability you want. They may be embedded with a product such as Lexmark’s, or a CAD system. The first step is to decide where and for whom you need search to be present.

Time to re-check your mobile development strategy

The mobile platform landscape has changed dramatically in the last few months. So much so that organizations who even recently reached decisions on a mobile development strategy should re-visit their decisions. I’m not talking about HTML5 vs app development issues – though those decisions are just as important and directly related because of continued innovation in device and operating system capabilities combined with the need to protect content development and management investments – but about which platforms will be viable, or meet your level of risk tolerance.

What has changed? To over simplify: Apple’s dominance continues to increase and is unassailable in tablets; RIM is not a contender; Microsoft is looking like an up-and-comer; and most surprising to many, Android is looking iffy and is a flop in tablets with the exception of the very Amazon-ized version in the Kindle Fire. These are pretty general statements, but if you are in charge of your company’s mobile development strategy considering their impact is a good place to start a check-up for a possible course correction.

Another place to start is to read the excellent post by Tim Bajarin Why Google Will Use Motorola To Become Vertically Integrated. I won’t summarize because the entire post and the comments are really a must-read.

Making big data analytics accessible to marketers

The recent announcement of SAS Visual Analytics highlights four important characteristics of big data that are key to the ability of marketing organizations to use big analytic data effectively:

  • Visualization is a challenge for big data analysis and we’ll continue to see new approaches to presenting and interacting with it. Better visualization tools are necessary not just because those who aren’t data scientists need to understand and work with the data, but because the increased efficiency and time-to-reaction to the data is critical in many cases – especially for marketers who need to react with lightening speed to current user experiences.
  • In case it isn’t obvious, visualization tools need to work where marketers can access them on web and mobile platforms.
  • In-memory data processing is necessary to support the required speed of analysis. This is still rare.
  • Big data is not only about unstructured data. Relational data and database tools are still important for incorporating structured data.

SAS is far from the only company driving new big data analytic technology, but they are the biggest and seem determined to stay on the front edge.

« Older posts Newer posts »

© 2024 The Gilbane Advisor

Theme by Anders NorenUp ↑