Text Analytics SummitPolySpot: Agile Enterprise Search Infrastructure

Cablesearch: Well, This Will Get Some Attention

December 8, 2010

Short honk: One of my handful of readers sent me a link to a search engine for Wikileaks’ cables. I have seen a cable for a long time. In fact, the last time I saw a cable, I was getting a demo of a system that allowed cable operators to see horoscopes and  lists of top 40 songs. Not yesterday exactly. More like 1974.

If you want to search Wikileaks’ cables, point your browser thingy at http://cablesearch.org. I don’t know how long this service will be online. I don’t want to hazard a guess. Lots of folks are agitated about Wikileaks so the goose will remain in Harrod’s Creek and maintain his interest in search technology. Seems like a wise move.

Stephen E Arnold, December 8, 2010

Freebie

Business Intelligence Bubbling

December 8, 2010

Open Source Business Intelligence: What Are Your Options?” investigates the momentum of OSBI in recent years.  In the words of Joe Nicholson, marketing VP at Pentaho, “The biggest trend we see is the rapidly increasing adoption of OSBI as a means to provide critical reports, dashboards and analysis to business users at far less costs than the traditional, proprietary BI solutions”.  It appears the interest in OSBI has been on the rise for quite some time, though what this will mean for consumers remains unclear.

What we are seeing is a good deal of posturing from the leaders in the field. With companies like Jaspersoft, SpagoBI and the aforementioned Pentaho, which claims it saves clients $1.5 million in fees over a three year span, clamoring for a spot at the head of the queue speculation on the future of the OSBI market is dramatic.  Aside from lowering costs, the idea of tapping into innovations spawned from the OS community itself has taken root, which according to Nicholson translates to more than six million downloads, upwards of eight thousand active projects and more than twelve hundred commercial customers.

Currently the community versions are available in a free downloadable format though they lack certain administrative and security functions, in addition to diminished ability to connect to prime data outlets. One still has to pay for premium access.

Sarah Rogers, December 8, 2010

China Nervous about Costs Too?

December 8, 2010

China Daily reports in the article Increasing Costs Worry Business Managers that China is facing its own economic woes. A survey was conducted with Chinese entrepreneurs to calculate their biggest challenges.

“The survey, conducted by China Entrepreneurs Survey System (CESS), found that 72.5 percent of the Chinese entrepreneurs chose “rising costs of labor” as the biggest difficulty in operating their business, 12.8 percentage points higher than the figure for last year. Also, the rising cost of raw materials was singled out by 56 percent of the Chinese entrepreneurs, compared with 48.5 percent in 2009.”

Despite the growing costs, product sales prices have remained stable and

entrepreneurs saw an increase in their profits at the beginning of 2010. China appears to be recovering from the economic crisis that has rocked the entire world. Entrepreneurs are optimistic about the future, but they are still leery, and are searching for ways to improve/upgrade their industries.

Whitney Grace, December 8, 2010

Freebie

Storeslider Search

December 8, 2010

StoreSlider is the new, more chic, way to search on eBay.

StoreSlider utilizes thumbnail size pictures in order to give consumers a faster, more convenient search experience that is also visually appealing. The thumbnail includes valuable information like the name of the store, when the auction ends and the current bid.

I know not everyone is a “visual” person, but if you happen to be like me; pictures are a good thing.

EBay’s StoreSlider is creating a more realistic and less tedious shopping experience by allowing its consumers to do the same thing they would do if they found something they disliked at the mall; skip it and move on.

For the more traditional shoppers, eBay will still allow you to click on the picture and see the actual auction information where you identify the normal eBay spread and cast your bid.

It seems that eBay is really tapping into the consumer need to “grab and go” by creating StoreSlider. Interesting but this UX or user experience approach does little to change the underlying functionality of eBay search.

Stephen E Arnold, December 8, 2010

Freebie

Mid Tier Consulting Firm Explains Enterprise Search

December 8, 2010

I have had a tough couple of weeks. My top engineer, Tyson, died suddenly. I had to attend a couple of meetings. I sat with a group of so-called search experts for 30 minutes before I ran in fear from a Starbuck’s in Washington, DC. Before I bailed from this intellectual meeting of the officers of the HMS Search Titanic, I flipped through a document from a US mid tier consulting firm.

Some ground rules. In this type of blog post I cannot name the analyst who wrote the document I scanned as I guzzled hot chocolate and gobbled a banana nut muffin. I wish I could identify the consulting firm, but I don’t need their 28 year old in-house lawyer telling me that I can’t use the name of the firm, mention its intellectual Mt. Zion, or suggest that the company has folks on staff who could neither get a job at a blue chip consulting firm or at Wal*Mart as a greeter. You will have to read my comments and do some thinking. Finally, I don’t want to reveal the names of the search vendors who got plastic laurel leaves nor can I name the vendors who got kitchen duty. Tough. Don’t forget to check out the editorial policy by clicking the About tab.

For those not familiar with my terminology, a “mid tier consultant” (sometimes called an azure chip consultant or an azurini as a reminder that these bench warmers are not the first team for the Bain, BCG, Booz, or McKinsey type of consulting service). I know the azurini can use an ATM or an iPads. I just don’t want these folks recommending mission critical enterprise software.

From the green flat the document stalled. I thought it was a spoof by Mad Magazine or the Onion. The write up was 10 pages long, had no page numbers, and not much accurate information in my opinion. The topic of the write up was “enterprise search”, a field which I pronounced dead after slogging through the first three editions of the enterprise search report between 2003 or 2004 and late 2006.

Two points. The firm responsible for this report recently deep sixed a four compartment analysis of enterprise search. Second, the new effort kicks off with a 30 plus word sentence with pontification, buzzwords, and few definitions. What is “information access technology”?

Enterprise search is dead, but it has been reborn in the form of several specific products and services. These include, but are not limited to:

  • Open source search options
  • Data fusion solutions
  • Search enabled solutions
  • Vertical search solutions
  • Niche solutions
  • Baked in solutions.

Pretty complicated? Yes, and getting more complicated. Search and retrieval is not simple in a technical sense. Even the most basic key word system can give system administrators and users a headache. The more sophisticated information access methods rank among the most challenging problems in computer science. Simple is easy for a marketer to say. But a Madison Avenue copywriter cannot change the difficulty of getting a particular user the information he or she needs in a way that is reliable, affordable, reliable, and accurate.

I have a list of enterprise search vendors that failed. The list includes Convera, Delphes, Entopia, and a number of others on an endangered species list. Why did these vendors fail? Why are some vendors

  • Search vendors for the most part have hit a “glass ceiling”. One example is a firm that generates most of its revenue from eCommerce and is trying like the beavers in my goose pond to gin up money from business intelligence sales. The reason I mention this is that the company in question had to get an intake of cash a couple of years ago, has been unable to find a buyer among those who invested in the company, and makes sales by enlisting the hottest academic business poobah from the greenest of the Ivy League. The only problem is that this particular vendors does not handle structured and unstructured data particularly well, suffers some performance problems when index cycles squeeze toward near real time, and charges a lot of money to address certain problems that have persisted since the late 1990s. Imagine my surprise when this vendor received a gold star and a kiss on the cheek from the mid tier consulting firm’s write up.
  • Open source has become a big deal for several reasons. First, there are quite a few vendors offering open source software and engineering services. The idea is that open source search comes without hefty licensing fees or handcuffs. Among the vendors I track are Tesuji in Hungary and Italy, Lucid Imagination in Silicon Valley, FLAX in Cambridge, England, and dozens of other vendors who stuff open source search into other applications. One interesting example is Pentaho, and there are others. I want to include dear old IBM, but that outfit has engineers who struggle to post comments to this Web log. Even though OmniFind in its many guises is Lucene, and you get IBM’s version with the same profligacy of Lotus Notes. The open source search option is somewhat disruptive, and the legal hassle between the kind and gentler Oracle boss ( Larry Ellison) and the Math Club’s stand up comedian (Eric Schmidt) may shake up the open source search market in some interesting ways. Open source does not get too much attention in the write up from the mid-tier expert in my opinion. Maybe a mistake? Maybe a certain blindness? Maybe a neural discontinuity? Maybe an English major longing for a “red wheelbarrow” metaphor?
  • The “give away” approach to search and content processing is another important aspect of enterprise information management. Here’s what’s happening, although you wouldn’t know it from the white paper I just read. Outfits like Microsoft and Oracle, among others, often view an organization in terms of total revenue across products and services. If one of these outfits has a customer looking to buy 200 client access licenses or adding 200 connectors to a data management system, the vendor looks at the big picture. Tossing in software in order to get the license fees, the hardware sales, and the engineering billability makes perfect sense. The client, therefore, gets a search system and for many information technology professionals who find information retrieval an annoyance, this bundle makes sense. The problem is that the well known dissatisfaction with enterprise information access exists regardless of the vendor whose search system is being used. This “give away” approach and the dissatisfaction combine in fascinating ways. Not a peep, of course, in the white paper.
  • In basketball, coaches look for vertical leap. in enterprise information access, the action has shifted dramatically to vertical solutions. The idea is that boiling the ocean does not work and has not worked since STAIRS III, InQuire and Personal Library Software, among others, were available. We’re talking a long time, folks, not 15 years. The thriving vertical sectors reveal a wealth of innovations. Just today I learned about a new company’s data fusion approach to financial analysis. Is this “enterprise search”? You bet your boopie it is. The vendor has wisely decided to focus on a specific problem that has high value to a particular market segment. This is search with a PhD in streetology from the School of Hard Knocks. The mid tier firm’s freshman composition is blind to this important development.

What concerned me was that some of those in the Starbuck’s discussion believed what was in the consultant’s white paper. Even more troubling was the lack of intellectual curiosity about the information, the framing of the argument, and the criteria for including one firm and ignoring another.

The factual errors were plentiful. Here’s one example of a serious factual error. First, navigate to www.gsaadvantage.gov and run a query for the GB-7007. Look at the license cost for a single server than can index 10 million documents. Now look at the license cost of the back up server. Now calculate how much the license fees will be over a period of two years. If you are too lazy to do this work, then you are the ideal customer for the mid tier consulting firm’s services. If you do the work, you will see that the cost is in the millions of dollars. Remember you are looking at government discounted prices, so a commercial outfit will pay more. I cover this pricing information for the Google Search Appliance in my KMWorld column about Google Search Appliance pricing. Hey, why worry about facts? That’s what makes you easy pickings for the azure chip carnival workers.

I could tiptoe through the dead intellectual tulips in the white paper, but I am bored and I don’t think there are too many people who know or care much about “enterprise search”. The buzzwords and the baloney are what makes advisors get up in the morning and greet each day with a big smile.

I snagged a Metro and headed for the airport. Back to the goose pond, away from the IQ degrading discussion about the mid tier consulting firm’s periscope or endoscope or whatever.

Stephen E Arnold, December 8, 2010

Freebie

Basis Tech Gets In-Q-Tel Funding

December 7, 2010

Calling the CIA “The Company” doesn’t seem like so much of an ironic joke these days with government funding developing technology that might help spooks spy.  “How the C.I.A. Perfects its Social Media Monitoring Technologies” gives an overview of the workings of In-Q-Tel, the CIA’s 10-year-old venture capital company.

Kashmir Hill writes: “In-Q-Tel is a rather secretive group. It declined to speak to me for the story, but I did chat with the CEO of one of the companies the group has funded. Basis Technology CEO Carl Hoffman told me In-Q-Tel is a great investor for a small start-up because it’s a gateway to Washington for small companies that normally struggle to compete for federal contracts.“  Basis Technology has been the recipient of In-Q-Tel funding to expand its multilingual text analytic software products and services to include Middle Eastern languages.  This new niche has, in turn, led Basis Tech to do business with several other federal agencies, an exciting development.

Hill’s conclusion is that the only downside for these CIA-funded companies is that it might hinder their expansion to China and the Middle East.  With terrorism paranoia at the levels that it is these days, I think he might be right.

Alice Wasielewski, December 7, 2010

Freebie

Email: The Humpty Dumpty for 2011

December 7, 2010

Facebook’s announcement of Titan, its new messaging system, has provoked a lot of analysis of the future of email.  “Facebook Messages Won’t Replace Traditional E-Mail, Poll Says” reports that the Wall Street Journal found that “More than 62 percent of over 3,680 participants in a recent online poll said they wouldn’t use Facebook Messages as their primary e-mail service.”

My initial thought is that the WSJ readership probably does not accurately reflect the overall FB demographic.  And, of course, do people really know what they will use until they’ve used it?  Most people had no idea four years ago how ubiquitous social networking would become.  Cringely weighs in with “The Decline and Fall of E-mail,” where his main point is that the death of email is from spam and the rise of social networking.  I wonder if there won’t be a different type of spam, or something spam-like on Titan.  In “Distinction Between Email and Social Networks Eroding,” WebProNews reports: “The popularity of social networking services, along with changing demographics and work styles, will lead 20 percent of employees to use social networks as a main business communication tool by 2014, according to a new report from Gartner.”

Gartner sees collaboration moving to the cloud  with companies like Microsoft and Research in Motion (RIM) cashing in.  None of this comes as a huge shock to me.  I think even the Luddite-ish among us have replaced much of email with FB already (Somebody needs tell the WSJ!), and with the size of Facebook’s membership, any additional incentive to message there will cause use of traditional email to shrink.  I think email will always have it’s niche, but so does the postal service, radio, and printed magazines.

Alice Wasielewski, December 7, 2010

Freebie

Restaurant Engines

December 7, 2010

I am not a big user of restaurant search engines. Here in Harrod’s Creek we have River Creek Inn and a McDonald’s. No need to search, grip a Groupon coupon, or consult a mobile mapping service.

Restaurant Engines is a new full service search engine that offers restaurants a three tiered approach to advertising. Not only will they build you a full service, fully interactive, website; they will also create an iPhone and Android application that allows users to easily find directions, set up reservations, and order from the menus.

In theory this is a wonderful idea that has the potential to create mass revenue by utilizing inexpensive internet access. However, in reality, I don’t see it being that effective. I can already order a pizza and set up a reservation from my phone. This new service would be a convenience but not by much, it seems to be a replica of what is already available and unless they have a truly remarkable advertising scheme, traffic may be hard to come by.

Stephen E Arnold, December 7, 2010

Freebie

Wave Wades Splash to Open Source Shore

December 7, 2010

As reported by h-online.com, Google’s Wave communication platform, forsaken sometime in August of this year due to lackluster user interest, has been repackaged and is in some capacity available to the public beneath the Apache banner.  “Wave in a box”, its new moniker, is a scaled back version of the original offering customers the ability to construct their own Wave servers in an effort to jumpstart the products development.  Discussions continue in reference to adding APIs, as well as a Web 2.0 client.

Per the article, “The proposal was posted to the Apache Incubator wiki by Google and Novell employees, as well as several independent developers.  The Apache Incubator is the place where potential future Apache projects can be submitted to the open source organization for consideration.”

Despite the proposal stating the project’s direction, it appears that much of the project is still in the development stages.  As of yet the code base remains at Google with the stated goal being the integration of it to the Apache infrastructure sometime in the future.  There is also speculation as to what sort of vendor the revised software will attract, with names such as Novell and firms related to the U.S. Navy being dropped.

Sarah Rogers, December 7, 2010

Freebie

Digital Reasoning Unleashes Synthesys Version 3

December 6, 2010

Our sister publication covers the dynamic world of data fusion and next generation analytics. I wanted to call your attention of an interview with Tim Estes, the founder of Digital Reasoning. The company has announced a new version of the firm’s Synthesys product. You can read a complete, far ranging interview with Mr. Estes in the Search Wizards Speak series at this link. Our analyses of the Digital Reasoning technology are most encouraging.

Here’s a snippet of the interview’s contents from the Inteltrax story which ran earlier today:

Synthesys V3.0 provides a horizontally scalable solution for entity identification, resolution, and analysis from unstructured and structured data behind the firewall,” Estes said when asked about Digital Reasoning’s new offering. “Our customers are primarily in the defense and intelligence market at this point so we have focused on an architecture that is pure software and can run on a variety of server architectures.” In addition, the program is ripe with features that are miles beyond previous versions. “We’ve enhanced and improved the core language processing in dramatic ways. For example, there is more robustness against noisy and dirty data. And we have provided better analytics quality. We have also integrated fully with Hadoop for horizontal scale. We probably have one of the most flexible and scalable text processing architectures on the market today.”

While the company still works heavily with the government, Synthesys technology will benefit several other fields. “We are getting good bit of interest from companies that need what I call ‘big data analytics’ for financial services, legal eDiscovery, health care, and media tasks.” For example, the program: “can identify the who and the what, map the connections, and deliver the key insights.” Estes continues, “instead of clicking on links and scanning documents for information, Synthesys Version 3.0 moves the user from reading a ranked or filtered set of documents to a direct visual set of facts and relationships that are all linked back to the key contexts in documents or databases. One click and the user has the exact fact. Days and hours become minutes and seconds.”

Read more

« Previous PageNext Page »

  •  Only search links from this page: