Text Analytics SummitPolySpot: Agile Enterprise Search Infrastructure

BBC Journalists Show News Nose

November 7, 2010

Short honk: I am not a real journalist. I don’t have a pension. I don’t strike. “BBC News Staff Strike over Pensions” alerted me that real journalists write large with their actions. I am certain the rationale is solid. Nevertheless, I noted this passage as exemplary:

Members of the National Union of Journalists at the BBC are taking part in a two-day strike in a dispute over proposed changes to the pension plan….When employees draw their pension, payments will increase automatically each year in line with inflation, by up to 4% – again up from a previous offer of 2.5%. Bectu, which represents technical and production staff, said after last month’s ballot that the amended offer was “the best that can be achieved through negotiation”.

The impression this leaves upon me is that what is offered is not enough. The goose gets nothing yet does not complain. Real journalists do that I suppose. I do remember that the BBC search boss told me that the BBC had search nailed. I suppose that type of confidence extends to financial acumen as well. Pension funding and search are obviously no brainers.

Stephen E Arnold, November 7, 2010

Freebie, just like a pension except without the money part.

Our Bad, Says Google. Oh, No Dough for You

November 6, 2010

Google has settled a class action lawsuit that was brought against it because of privacy violations surrounding Buzz, the company announced last week. Google Buzz, the company’s latest attempt at a social product, was a disaster at launch because it made much private information public without any explicit user consent. In response to backlash, the company quickly introduced improvements. The settlement acknowledges the improvement attempts by Google but still calls the company to establish an $8.5 million fund “the majority of which will go to organizations focused on Internet privacy education and policy” according to “Google Settles over Buzz, Will Establish $8.5 Million Fund to Promote Privacy Education.”

Is $8.5 million the price of Google’s controlled chaos? No, we think it will continue to rise, especially since $8.5 million isn’t enough to raise eyebrows. And, no dough, for you.

Leah Moody, November 6, 2010

LexisNexis Adds Content

November 6, 2010

LexisNexis is bringing the world together one search at a time. “LexisNexis TotalPatent Becomes First Patent Research Service to Provide Direct Links to….” Practically everywhere! LexisNexis’s new version of TotalPatent provides researchers the ability to search SciVerse Scopus, a virtual plethora of articles just waiting to be utilized by professionals in the scientific, medical,technical and social science fields.

“In conversations and visits with our customers, we learned how valuable it would be for them to have direct access to non-patent journal articles without having to leave TotalPatent and conduct those searches independently,”

TotalPatent also links up to Chisum on Patents. Chisum is the foremost authority on patent laws in the United States. Here’s the really cool part. TotalPatent includes English language software. It can translate foreign articles into the English language. All of these new improvements allow researchers to work more efficiently and accurately.

No word on prices for the new service.

Leslie Radcliff, November 6, 2010

Freebie

Oracle and SAP Amusement Park Admission Set

November 6, 2010

SAP will pay Oracle $120 million for reasonable attorney’s fees and costs in the past and future under terms of a joint stipulation filed Monday, said “SAP Willing to Pay Oracle $120M for Attorney’s Fees.”

Oracle sued SAP (www.sap.com) in 2007 alleging that its TomorrowNow subsidiary “illegally downloaded Oracle software and support materials in order to lure away Oracle customers.” The lawsuit is ongoing currently.

Under the agreement, TomorrowNow stipulates to entry of judgment on Oracle’s claims and in turn, Oracle (www.oracle.com) will not seek punitive damages against TomorrowNow or Sap. The trial jury has been instructed to only consider damages available under the Copyright Act.

The court must approve the agreement to make it binding, and either SAP or Oracle can withdraw if any provisions are not accepted.

Despite the elimination of punitive damages, SAP is positioned to give Oracle far more the claimed $2 billion damages by Oracle. Looks like $125 million was a low ball offer.

We think the real question is: What’s the price of happiness for Larry Ellison? That’s what’s coming out your pocket SAP.

Leah Moody, November 6, 2010

Freebie

Yellow Pages Turning into Ash?

November 5, 2010

The evergreen yellow pages seem to lose their identity amidst the multi-colored Google Places. Having amassed the information of about 50 million places, the global Google can now act local. The article Google Introduces “Place Search” reports that this new feature would serve the individual needs of users who search for their area-specific information, and benefit Google as stated:

From a business point of view, the localized content search might attract small-scale business to advertize on Google, as the target of such localized slides will be users from their area who represent potential customers.

Google bases this new local search feature on the fact that most search requests made are concerned with a location-search. The article “Google Launches Smart Search for Places” apart from detailing on Google Place Search, informs that its mobile version is also in the offing. It also reflects the view that the mini pages of local information based on the Google search algorithm, slowly deems to doom the commonly used yellow pages.

Harleena Singh, November 5, 2010

Freebie

Misunderstanding Facebook?

November 5, 2010

I just read “Big Deal: Facebook emerges as Major Player in Mobile and Location-Based Services.” In a sense, I agree with the write up. On the other hand, I think that the article is one of those summer stock efforts. You know the play was written by Shakespeare, but what the heck happened to make Act II so confusing.

Here’s a passage that resonated with me:

As context for all of this the company said it has an active mobile user base of 200 million people (out of more than 500 million total users). It doesn’t break out US vs. non-US numbers — though I’d bet the majority are in North America. If even half of those users are in the US it would make Facebook as large as Verizon. The difference is that Facebook’s members are much more engaged.

There are two really important points embedded in this snippet of text, and I want to highlight those and show why Facebook is in a state of increasing misunderstanding among the azurini.

The two words:

Active
Member

These two words are a very big deal, even bigger than mobile. Here’s my view.

image

Facebook creates a perception of this type of environment. This is not like using the services provided from the local electric or water company.

Word one: Active. As a touch point, consider the Google. Google has lots of users. The users navigate to a Web page, do something, and hit the trail. I know that lots of Google users provide some information about themselves and that lots of Google users rely on various Google services. I rely on the electric and water company, but I don’t spend what my boss at Booz, Allen used to call “quality time.” Google and some other services are like plumbing. Essential—I don’t talk about it at lunch. Facebook users are active. Active means habitual access. Habits in online behavior are good. But a habit like using an electric light are hard to break, but they are still utilitarian functions. Part of the woodwork. Have you hugged your door frame today?

Read more

HP a Marketing Marvel

November 5, 2010

I found “HP’s Move on IBM” thought provoking for two reasons. First, the news angle that HP was going to put a mixed martial art choke hold on IBM was a surprise. IBM is an acquisition-focused services firm. The goal of IBM, in my opinion, is to sell high margin services. The notion of IBM building a product that sells like an iPad is not on my radar. Even in search, IBM defaulted to open source, not its own basket of information retrieval novelties.

The second reason was the phrase “Instant-On Enterprise.” True I worked for Halliburton and a couple of other not-so-interesting companies in my career. One thing I learned during that tenure is that not too much within an established organization is instant on or off for that matter. I suppose there are some executives who want the world to work just like a Google search, but the reality is that most procurements take time. Change is resisted, often in ways that Mohandas Karamchand Gandhi would find exemplary. Once an engineers gets certified for some enterprise solution, that engineer thinks, “Keep job.” Once someone figures out how to run a trade show, forget the notion of learning a new trick. A trail ride horse knows where the barn, hay and water  are. Ever try to get the trail ride horse to go someplace else, you—you digital Hopalong Cassidy? Yep, instant on.

Get that horse trained to do another trick, Hopalong. Stuck. Get help from Cisco.

Here’s a passage from the article that I marked as significant:

It is convincingly packaged, but comes with problematic timing. It is, HP officials have acknowledged, the product of strategic work by Mark Hurd, who was fired by HP’s board in August after some head-scratching indiscretions with a good-looking contract employee. Apotheker, a former head of enterprise software company SAP, is now ultimately responsible for the grand configuration of Hurd’s acquisition.

Okay.

I think HP can mosquito bite IBM. HP is not a predator that will keep IBM’s high margin consultants awake at night. Perhaps after Mr. Apotheker surfaces? Perhaps after the instant guy settles in at Oracle and gets a new idea? Perhaps when the HP way instantly clicks on? I keep in the back of my mind that HP once owned AltaVista.com and let those engineers drift down the river of time to the Google.

Stephen E Arnold, November 5, 2010

Freebie

Oracle and Open Source Revenue

November 5, 2010

Oracle Raises Cost of Low-End MySQL Support” signals the onset of toll booth construction for open source software on the information superhighway. The write up summarizes “low end” prices. If you are into $600 to $20,000 prices, and possibly higher, you will realize what the high rent district is like. Here’s a sentence noted:

Oracle’s apparent pricing changes may irritate some MySQL customers.

So what’s left unsaid? First, the notion that open source software is a real deal is likely to shift. Oracle is an alpha male and lots of followers go along for the ride.

Second, if open source prices rise, that may make bargain priced enterprise software vendors look more attractive. The benefits of open source are often price. When the cost of support makes the costs look alike, I know some folks who like the idea of full time engineers managing a software product from an established vendor. There are, for example, a number of search and content processing solutions that are reasonably priced now. If open source search support rises, those solutions—mostly from European software firms—are going to look better, much better in my opinion.

Third, Oracle is a company that must generate lots of revenue. Like a shark, the outfit has to find a way to keep money growing. One way to do that is to squeeze revenue from open source and be aggressive to any firm that gets in its way. Even Google is not exempt from Oracle’s desire to get paid for its “open source” properties.

Such notions do not creep into ComputerWorld. How open source is open source?

Stephen E Arnold, November 5, 2010

Freebie

IBM OmniFind Text Search Server White Paper

November 4, 2010

IBM has published “Exploring the IBM OminiFind Test Search Server.” You can, as of November 3, 2010) download a copy without charge from this link. The page on the IBM site for this publication is on the Partnerworld subsite. The white paper covers OmniFind (based on Lucene) as a system that performs “high speed linguistic text searches against DB2 text data and documents stored in rich-text formats.” This suggested to me that OmniFind can query supported files in structured and unstru8ctured formats. Structured, as I understand the term, means content residing in relational databases like IBM’s own DB2. Unstructured, as I understand the term, means information like email, standard office worker file types like Microsoft Word, and Web pages. A list of supported file types appears in the white paper, and it includes a representative sample of what OmniFind can access; for example, XML, the ever popular Lotus Freelance, JustSystems Ichitaro, and Quattro Pro. The white paper dives into specific commands that are useful to an engineer installing the system. One interesting feature is that the system makes it easier to index “external data”. Manual inspection to make sure what you want indexed is indexed is a useful best practice. The XML search discussion makes clear that the user should be comfortable looking at XML mark up. Bob and Betty in marketing are likely to find the tags off-putting. If you are interested is learning how open source search technology can be used by a proprietary company in a commercial product. The white paper is worth a read and I think it is better than the writes up from some of azurini, who have discovered revenue by selling marketing services to search vendors.

Stephen E Arnold, November 4, 2010

Freebie

Google and Artificial Intelligence

November 4, 2010

Short honk: For those interested in how Google uses “artificial intelligence”, there’s a new patent that I found interesting: “Automatic Generation of Rewrite Rules for URLs” (US7827254), filed in December 2003, when Google was at the outset of its phenomenal run up to its peak in 2006-2007 in terms of technical productivity. You can download this document from www.uspto.gov. Make sure you work through the query syntax. The USPTO’s system is quite a delight to use. One of the inventors is Peter Norvig, yep, that Dr. Norvig. His book on artificial intelligence is one of the most thumbed in Math Clubs worldwide. The claims are interesting as well. Some are broad little devils. I thought, “Maybe Bit.ly” could be renamed “Bitten.ly”.

Stephen E Arnold, November 4, 2010

Freebie

« Previous PageNext Page »

  •  Only search links from this page: