Interesting Google Spoof
March 10, 2011
Short honk: We are not sure how long this system will remain online. You can see the GOPgle service at http://uxux.org/gopgle/. When you run a query, the system generates a results list that ranks the best or the worst of a certain political party. The system may be dead by the time you read this. Clever in a way while it lasts.
Stephen E Arnold, March 10, 2011
Freebie
Rosette Linguistics Platform Releases Latest Version
March 10, 2011
Basis Technology has announced its most recent release if its Rosette Linguistics Platform. Rosette is the firm’s multilingual text analytics software. Among the features of the new release is the addition of Finnish, Hebrew, Thai, and Turkish to the system’s 24 language capability. One point that we noted is that this release of Rosette sports an interesting mix of compatible search engines. According to the Basis Tech announcement:
“Bundled connectors enable applications built with Apache Lucene, Apache Solr, dtSearch Text Retrieval Engine, and LucidWorks Enterprise to incorporate advanced linguistic capabilities, including document language identification, multilingual search, entity extraction, and entity resolution.”
Several observations seem warranted. First, Basis Tech is moving beyond providing linguistic functionality. The company is pushing into text analytics and search. Second, Basis Tech is supporting commercial and open source search systems; namely, the SharePoint centric dtSearch and the Lucid Imagination’s open source solution.
The question becomes, “What is the business trajectory of Basis Tech? Will it become a competitor to the vendors with which the company has worked for many years? Will it morph into a new type of linguistic-centric analytics firm?” Stay tuned.
Cynthia Murrell, March 10, 2011
Freebie
Patterns in Web Content
March 10, 2011
Data mining refers to a form of application which seeks common themes or patterns in specific pools of information. The core of its popularity rests within the scientific communities, though the technology is increasingly being applied in the various arteries of the commercial sector.
The exponential growth of the Web has brought into focus the necessity for the ability to trace and scrutinize the relationships inherent in the aforementioned collections of information.
The Computational Linguistic & Psycholinguistics Research Center (CLiPS) located in Belgium has just released Pattern, a mining unit that was designed to couple with the Python language system. The Pattern Web site says:
“It [Pattern] bundles tools for data retrieval (Google + Twitter + Wikipedia API, Web spider, HTML DOM parser), text analysis (rule-based shallow parser, WordNet interface, syntactical + semantical n-gram search algorithm, tf-idf + cosine similarity + LSA metrics) and data visualization (graph networks).”
When you follow the link above, you can access the release directly. Check out the the specifications for compatibility.
I thought it interesting to discover the designers, in a trial of their creation, used the software to track the progress of a local politicians in the 2010 elections in their home country. Pattern scanned thousands of Tweets, split between two languages, updating the data pool on a daily basis. The results were fascinating. You can read a detailed description of the experiment here.
Micheal Cory, March 10, 2011
Are Google Rules Made to Be Broken?
March 9, 2011
I am not into financial comparisons in the UK or in the US for that matter. Too many opportunities to fiddle the data exist. Your view of these types of online services may be different from mine. After all, I am 66 and have worked in the commercial online sector for 30 plus years. Yep, lots of looseness exists. Anyone remember the ratings of certain financial instruments coincident with the financial meltdown in 2008?
The story that caught my attention was “Google Buys BeatThatQuote, a UK Comparison Site Violating Google’s Guidelines.” The main idea was that Google acquired a semi-hot company for pocket change; that is, $60 million or so. The SEOBook article asserted:
What is screwed up about this is that Google is engaging in *major* channel conflict. Not only is there some gray area background stuff:
“BeatThatQuote.com’s ad prompted 101 complaints to the Advertising Standards Authority, with 65 objecting that the commercial “trivialized, condoned or encouraged bullying in the workplace”.”
But now they have to consider SEO as well. I highlighted how it was a bit unjust when Google arbitrarily chose to whack one site while letting another get away with worse just because the founder was good at public relations, but how can Google police Google’s guidelines when Google is the one breaking them?
Now this is quite interesting. Are Google’s rules made to be broken? Heck, when you are a really big outfit operating in post crash America with a great deal of cash, rules are mostly “suggestions” or “hints”.
I think that as consolidation continues in the online space, the spirit of AT&T before Judge Green will influence a number of firms’ acquisition activities. Google is not doing anything much different from some other firms in similar positions. The free market encourages rules that work much like suggestions or hints.
Adaptation is an important attribute.
Stephen E Arnold, March 9, 2011
Freebie
Recommind and End to End eDiscovery
March 9, 2011
“Recommind Advances End-to-End eDiscovery” explores the value of an eDiscovery implementation that relies on a single vendor’s solution. The idea is that an organization like Cisco can standardize on Recommind’s solution and reduce costs and get a boost in efficiency.
This is an interesting idea and one that has fueled many enterprise software vendors over the years. The key point of the write up, in my opinion, was:
eDiscovery…is a term that I have seen used both narrowly and broadly. Craig [Carpenter, Recommind executive] said it was initially used in a more narrow sense to refer to the process of preserving, collecting, reviewing, analyzing and producing relevant documents after a legal process has begun. Now, it generally extends to include information management in anticipation of possible legal issues, as well as during them, while also addressing regulatory needs. This reflects a more proactive approach that is being adopted by many Recommind customers. This broader approach means that content collection is not an isolated activity but one that is integrated into an enterprise’s total information management system on an ongoing basis.
This point was followed by this passage:
Cisco is using Axcelerate to take a more proactive, strategic approach to eDiscovery. One of the drivers of increased eDiscovery costs is the explosion of enterprise content, which ironically Cisco is helping facilitate with its technology that simplifies the creation and transfer of large files, such as videos. Cisco wanted to have its own enormous stores of enterprise data easily accessible ahead of any legal issues that might arise. They also wanted greater cost containment and predictability so that legal costs do not dictate when they settle and when they fight a legal challenge. Axcelerate offers customers like Cisco a comprehensive, end-to-end platform for eDiscovery management…This greatly reduces the amount of expensive manual labor by legal experts as they decide which documents are relevant and/or privileged in a case.
We think this is a good idea and a solid example. However, in many firms there are numerous legal matters underway at any one time. Not all firms operate from a single geographical location or rely on a single law firm. As a result, some organizations have no choice but use a variety of eDiscovery systems. Costs are important but there are other considerations as well, including the time available to the legal team and the preferences of outside counsel.
One other point is that there is often more to eDiscovery than processing digital content. Companies like Brainware and ZyLAB have gained traction with their ability to manipulate hard copy documents. Both of these companies assert an end to end capability as well. To be fair, Recommind may be focusing more on a narrower definition of the work flows in a legal matter.
We find the positioning of the firm’s technology interesting and will continue to monitor the firm’s description of its technical and marketing approach to a big and growing market.
Stephen E Arnold, March 9, 2011
Freebie
Lucid Imagination and a German Open Source Summit
March 9, 2011
In case you weren’t able to make it to Lucid Imagination’s recent 2011 Solr meet-up in Germany, the company has posted a couple of presentations here. We think you will want to download these. Our favorite is Marc Krellenstein’s but you should make your own determination based on your information needs.
The offerings include a piece on search trends by Marc Krellenstein, one of Lucid’s founders. He looks at such factors as evolving challenges, best practices, and the “current competitive landscape.” He also takes a look at Google as compared to enterprise search, and weighs the strengths and weaknesses of his company’s own Lucene/Solr.
We also liked Thomas Kwiatkowski’s and Oliver Schönherr’s discussion of the online real estate site Immobilien Scout. They detailed their methodical transfer from their legacy search and data infrastructure to Solr.
You will want to check the Lucid download location. When we tried to snag the presentation form Olaf Zschiedrich, head of German-language eBay, (eBay Kleinanzeigen), it was unavailable.
Lucene/Solr continue to gain traction in the Web and enterprise search market. For more information about Lucid Imagination, navigate to www.lucidimagination.com.
Cynthia Murrell, March 9, 2011
Is 97 Percent of a Market a Monopoly?
March 9, 2011
Quite a question which I have not seen anyone ask. I did read the startling spider food headline “Google Controls 97 Percent Of Mobile Paid Search: Report.” I wonder if any eagle eyed regulators paid attention. Now the data come from a source that may not be familiar to some—an analyst report from a financial outfit. Most of these data are carefully screened and often support a position that the analyst wants to take with regard to a particular firm or market sector.
Notice that the rotini noodles are 98 percent the same. Is this consistency or a monopoly among the rotinis?
Google is the outfit the possible monopoly position. I suppose that someone at Google was surprised to be given such a accolade. Google has worked hard to present itself as just another friendly competition. Now an analyst firm asserts that the data from Efficient Frontier “proves” Google is the big dog in mobile paid search.
Now I am not sure what “mobile paid search” is. When I looked for a Pizza Hut last night, I had to resort to calling someone to look up the phone number. I then called the Pizza Hut, ordered the rotini my 89 year old father wanted, and drove to the aforementioned Pizza Hut. I want to note that my trust BlackBerry map did not get me to the Pizza Hut. I drove around until I spotted the building tucked next to a $5 haircut outfit.
Here’s the key passage from the write up which is probably going to be cited a number of times in the next 24 hours:
Just as Google dominates mobile search share in the US (with roughly 98 percent), the report said that 97 percent of mobile search spend (for Efficient Frontier clients) now goes to Google, while 3.2 percent spend goes to Bing/Yahoo.
Not too many qualifiers in this write up. Check out the charts. The pie chart is a keeper and will probably surface at some Congressional hearing later this year.
Stephen E Arnold, March 9, 2011
Freebie
Protected: Microsoft Fast Context Properties
March 9, 2011
Is eBay Changing Direction?
March 8, 2011
Exorbyte just released some interesting news on its blog, “eBay is Magento’s Secret Investor – Internet Retailer”. It appears the leader of online marketplaces is finally catching on to its smaller merchants’ complaints. So much so that eBay had acquired a 49 percent stake in Magento, the open source ecommerce Web application. Exotbyte Commerce Search is available as a plug-in for Magento.
Here’s one snippet from the Exorbyte write up:
“This is confirmation that there will be an ongoing market of small online retailers who do not want to operate within the restrictive and expensive platforms of eBay.com or Amazon.com; where fees are high and they have no or little control of the customers relationships. This market of small online retailers using installed or hosted ecommerce platforms is where Exorbyte Commerce operates.”
The question in Harrod’s Creek is if eBay’s obvious need to appeal to its lower volume patrons foreshadows some version of a buyout on the horizon. We shall see. eBay has become dependent for sizzle on PayPal. eBay’s original service seems a trifle dowdy. Magneto is a fresh name at least.
Sarah Rogers, March 8, 2011
Freebie
IBM, OmniFind, Yahoo: What Is Next in Search?
March 8, 2011
I was pointed to this link from IBM’s website, stating that effective September 22, 2010 the Yahoo! Edition of OmniFind was withdrawn from marketing. Further, after June 30, 2011 the product support line will be nullified as well. Why the cancellation? There is an announcement letter attached, though it offers no insight. It does, however, indicate that no replacements will be made available. I tried tracking down some additional information on this free version of OmniFind and its demise, to no avail. So what happened? Was it not popular enough, too limited? Was it Lucene? Does anyone care? We do find IBM quite fascinating for a $100 billion company on Jeopardy, fighting crime in New York, and cracking medical problems with Nuance. Diverse for sure.
Our view is that with Microsoft getting cozy or at least semi-cozy with Yahoo, IBM’s OmniFind unit saw the writing on the wall. The result is that the “free” and severely limited IBM Yahoo search solution was pulled. What will take its place? Our first thought was Watson. Then we realized that Watson may be more appropriate for limited domain searching like medical information. The Jeopardy public relations stunt was a marketing play. Configuring Watson to replace Lucene based OmniFind is probably not a practical solution at this time. As a result, IBM will have to either create a new demo of its enterprise search software or just roll out the full system with some limit on its use.
In short, IBM like other big outfits seems to have some challenges in the enterprise search department just as Google, Microsoft, and Oracle do. No wonder the third party solutions are probing the enterprise for licensing deals. Stay tuned.
Sarah Rogers, March 8, 2011
Freebie
-
- Subscribe to Beyond Search
Feature archive
News archive
-
