SLI Receives Big Boost From Trade Me Founder: Morgan Invests Million
February 20, 2013
I read an interesting article this week about SLI Systems. It seems the full service search and navigation based company has generated some interest from a wealthy benefactor.
According to Business Day’s article “Morgan Invests in SLI Systems,” Trade-Me Founder Sam Morgan has invested $1 million in new equity in the provider. This is good for SLI because the equity can be used to inflate employee numbers and create growth within the company.
“Morgan said the potential for growth in global on-line sales made SLI Systems an exciting investment. “Online sales are such an important segment for all retailers now and SLI simply helps customers find things and buy them more easily.”
It’s that potential growth that has Morgan interested in the investment and he has been involved in a number of other business investments that promise to make small businesses run more smoothly. Most recently he’s invested in Xero, the online accounting software.
Morgan has been able to take Trade Me from a one man operation to the largest internet auction website in New Zealand but can his investment and influence take SLI to the next level? It remains to be seen, however, the money will serve to grow the number of employees and market reach simply because of the Morgan name.
Leslie Radcliff, February 20, 2013
Sponsored by ArnoldIT.com, developer of Augmentext
Facebook Imagines A Whole New Way To Search
February 19, 2013
According to “How Facebook Made Me Search In Strange New Ways” from Search Engine Land, Facebook Graph Search has changed the way people search or at least wants to. Before social networking, people used to actively search with search engines because an action influenced them. With social networking, people are passively discovering information that they never knew they wanted to know all due to network connections.
Graph Search will not replace Google, because it does not search the entire Web. It only contains social information, in other words a social search:
““Social search is a means of uncovering information, to fill a knowledge gap taking into account crowd sourced information from your network, which contains information from a reputable source within your network, giving more of a credible touch to the content.”
Graph Search does not have all of this, but with some improvements it can. It can get better by adding general searching, i.e. only search through Facebook while Bing picks up the regular Web and the social network. The image search can use tweaking by searching tags and captions. Also the Facebook auto complete feature is annoying and is down right wrong sometimes.
Graph Search is not grasping the basic search engine fundamentals and needs to add them to even think about being competitive. The “graph” will probably be the next wave in social networking, so it is necessary to get on it.
Whitney Grace, February 19, 2013
Sponsored by ArnoldIT.com, developer of Beyond Search
NIST Trec Results
February 18, 2013
Short honk: A person heard me give a talk in which I said, “NIST Trec results are not widely available.” I thought the comment was appropriate, but someone thought I was full of goose feathers. I want to point you to the Trec page at http://trec.nist.gov/results.html. On that page one can view the results provided the user has a valid user name and password. In my book, that means “not widely available.” Your mileage may vary. Basic information about TREC is at http://en.wikipedia.org/wiki/Text_Retrieval_Conference, which like most information about search is sort of right and sort of not right.
My hunch about the closed nature of the comprehensive data Trec results spit out is that some outfits participating use math magic to demonstrate a particular system’s performance. My inner goose thinks that the flat lining of precision and recall scores and the data about search and retrieval within certain test sets indicates that search is not making the sort of progress most vendors wish to be true.
Stephen E Arnold, February 18, 2013
Searching a Priority? Also a Source of Fascinating Confusion?
February 18, 2013
The headline of a news release from Concept Searching caught my eye: Survey Results Indicate Enterprise Search is still the Number One Priority.
The subtitle puts the data in what I call the “marketing context.” Here is is:
Recent Concept Searching Survey Finds CIOs are Still Struggling with Search
The basic findings are revealed in this paragraph:
With the explosion of unstructured content, enterprise search has become inadequate for the knowledge worker but also renders relevant content inaccessible when applying it to collaboration and text analytics. Organizations are recognizing that search is not a stand-alone technology or application, but must be integrated with business processes and corporate objectives as a key infrastructure component. Regardless of the enterprise search solution, the delivery of meaningful results depends on the ability to effectively index and classify content, and to develop taxonomies to better manage the content. With the unabated growth of unstructured content and the introduction of diverse environments, such as on-premise, cloud, or hybrid, the need to access relevant information is falling far behind organizational needs. Content can exist in multiple repositories, and if not effectively managed increases organizational risk.
The sponsor of the study is Concept Searching, and a “white paper” providing more information is available from the company at www.conceptsearching.com.
Several observations:
- Search is a nebulous term. Many vendors are using the nomenclature of customer support, data analytics, eDiscovery, and Big Data. The reason, based on what little I know, is that search is not a hot term in some circles. See, for instance, the “comScore Releases the 2013 Digital Future in Focus Report.” The message there is that the digital landscape is rich and varies, in flux, and moving in unexpected directions.
- Open source plays are picking up steam. Based on my recent interview with Miles Kehoe and a forthcoming interview with Mark Bennett, both at LucidWorks, organizations are embracing open source solutions for a number of reasons. Could this be good news for open source solutions and less good news for vendors of walled garden systems which support “standards”? With the wave of consolidations and the HP Autonomy dust up, has cold water been tossed on the raging fires of information retrieval?
- Good, bad, or indifferent, Google has become the touchstone for many people when the word “search” is used. I saw a discussion on the LinkedIn search forum which presented the idea that appliances are back in vogue, specifically, the Google appliance. Simplicity may be gaining traction in markets confused or under cost pressure. Do licensees want complexity, multiple moving parts, and highly refined components or a solution that works?
- Companies in data management like MarkLogic are going back to their roots in publishing because the search push has not resulted in the revenue windfall some anticipated. Have the hopes for big paydays via enterprise search shifted to paydays for services firms and consultants?
Search, it seems, remains fascinating and confusing. Thank goodness I am old and content with my paper files and steam powered computer.
Stephen E Arnold, February 18, 2013
An Overview Of eDiscovery From A Lady
February 18, 2013
The practice of e-Discovery is currently blossoming in the legal system. It provides important insights in litigation and evidence for cases, but what does the future have in store for e-Discovery? Laura Zubulake was a plaintiff in a case that changed the legal system and how it uses technology. She was recently interviewed for Clearwell Systems e-Discovery 2.0 blog in the post, “Q&A With Allison Walton Of Symantec And Laura Zubulake, Author Of Zubulake’s e-Discovery: The Untold Story Of My Quest For Justice.”
Zubulake voiced that technology had advanced greatly for e-Discovery, but organizations have not made any great strides in information management. She believes as time goes on e-Discovery will become absorbed in information governance, though a need for e-Discovery experts will arise and require companies to bring in third-parties. Organizations will come to realize the importance of e-Discovery and start restructuring to include it. The government will be one of the main forces that drive e-Discovery adoption, especially when it comes to procuring information or evidence. State and federal governments will be among the first to have laws and procedures for using it. One of the biggest changes is that users will finally understand the power of search algorithms:
“Algorithms become an accepted search tool. Although keyword, concept, cluster, etc. searches will still play a role. For years, law enforcement, government, and Wall Street have used algorithms—the concept is not new and not without peril (significant market corrections were the result of algorithms gone wrong). Parties confronted with volumes of data and limited resources will have no choice but to agree to computer assistance. However, negative perceptions and concerns about algorithms will only change when there is a case where the parties initiate and voluntarily agree to their use.”
Will 2013 be the year e-Discovery is magically accepted without question? The technology will continue to advance and newer, “younger” organizations will be quick to adopt, but moving established organizations (like the government) is going to be pushing a boulder up hill. People like Zubulake are taking a hammer and finding the boulder’s weak point, then carrying the pieces up hill.
Whitney Grace, February 18, 2013
Sponsored by ArnoldIT.com, developer of Beyond Search
Experts Are Only A Search Away
February 18, 2013
Have you ever heard of Funnelback? Probably not, unless you are a search expert or come from the land down under. While the search experts are at work, allow me to explain a bit more about Funnelback. It is an enterprise and Web site search that sports an algorithm that can be tweaked to reflect a user’s customizable search results, editable search parameters, and a development platform for multimedia, e-commerce, e-mail media alerts, and plagiarism detection systems. The last option is one of the reasons Funnelback has gained a huge following in Australian universities.
If you are searching for a prime Funnelback experience, check out the University of Melbourne’s Web site with its “Find An Expert” search engine. The Funnelback “Find An Expert” searches through the university’s staff and faculty directory and retrieves experts related to the user’s keyword. In our example, we searched for “politics” and the results yielded 232 experts. Department, types of politics, and topics on politics, can filter the results. What makes Funnelback more entertaining than Google is that it creates a “capability map” aka a visual representation of the search results and how they connect with each other. The capability can be manipulated by filtering out or including other results.
Funnelback demonstrates that search can be entertaining and intelligent. When will Google add this to their search results page?
Whitney Grace, February 18, 2013
Sponsored by ArnoldIT.com, developer of Beyond Search
IBM and Price Cuts: Is Watson a Factor?
February 17, 2013
I read “IBM Cuts Price of Watson Based Power Servers.” I have no clue if the story is correct, half current, or incorrect. What’s important is that CIOL.com thought the notion of a Watson related price cut newsworthy.
The Power7 based servers were hot stuff several years ago. CPU performance is no longer the gating factor as it was in the days of STAIRS III. Input output, memory subsystems, and various types of latency make a system fast or not. Heck, careless programming can make Google’s zippy boxes howl with pain when its innards suffer a computational cramp.
The write up asserts:
IBM will roll out eight new Power Systems for entry level starting at $5,947. The new systems include Power Express 710, 720, 730 and 740 family of products…. IBM will also introduce two new PowerLinux Systems – 7R1 and 7R2 – optimized for IBM InfoSphere BigInsights and InfoSphere Streams big data analytics software. The company will also introduce two new Power Systems – 750 and 760 – for midsized and large enterprises.
The hot item in the story in my opinion is this reference:
The new systems are based on IBM’s Watson system and are powered by its Power7+ microprocessor technology. These will enable users to build and deploy infrastructure for private and hybrid clouds, as per a release.
The write up includes the now obligatory baloney about the big data, cloud and caching tactics for performance.
If the story is incorrect, no big deal. Any publicity is good, even for a dog movie like “Heaven’s Gate” and its expensive roller skates. If the story is half correct, why is Watson making an appearance in juxtaposition to “entry level.” Is the vaunted Jeopardy winning technology not generating sufficient revenue to payback the development time and the sunk marketing costs? If the story is correct, I am interested in the fact that high end information technology has to be bundled at lower prices.
Years ago, I was told by an informed person that IBM knew what it was doing when it came to search and information retrieval. Maybe the company will come to dominate the enterprise market for big data, analytics, and smarter search. On the other hand, hasn’t IBM travelled this road before and yet the journey continues.
Stay tuned to Jeopardy or monitor the cancer related news stream. Watson is with us along with a Power7 chip which may be experiencing some symptoms of rheumatism.
Stephen E Arnold, February 17, 2013
Facebook, Search, and the Real World
February 16, 2013
I think there is or was a television program about the “real world.” I am hazy on this, but I perceive “reality television” as a semi scripted, low cost way to fill the gaping maw of 24×7 programming at a bargain basement price. In fact, when anyone suggests that something is “real” I take a second look. This applies to “real stories”, “real life examples”, and “real consulting insights”. In today’s world, the notion of “real” is slippery. I think of Plato, Hollywood special effects, and marketing baloney.
I read “You’re Not Gonna Like It: Facebook’s New Search Struggles with the Real World.” The title caught my attention because of its use of the familiar “you,” the word “gonna”, the inclusion of “search”, and the phrase “real world.” In a horse race there is a big payday from picking win, place, and show. Here the headline snags the top four spots in the social media World Cup.
The article points out that some of the features of Facebook search need to be rethought. That’s a fair statement. The product is a beta and represents the first somewhat edible fruit of the marriage between the Facebook crowd, the injected Googlers, and the post IPO attention of the kind and loving stakeholders.
Facebook has to produce revenues, keep its costs under control, and cope with a number of darned exciting issues. These include the mandatory registration Google has slapped on Google Plus and the awareness by some Facebookers that there may be something else to do with the time invested in posting information about one’s comings and goings.
Here’s the passage I noted:
Facebook launched Graph Search at a big press event at its Menlo Park, CA headquarters almost exactly one month ago. CEO Mark Zuckerberg delivered a large part of the event keynote himself, highlighting the feature as one of three “pillars of Facebook” alongside the News Feed and Timeline. Graph Search is supposed to help you gather friends for a Twin Peaks marathon, find photos taken in London on your last trip, and see which sushi places are most popular among your friends. After a month of testing Graph Search, I’ve found that it’s fantastic at finding people and photos, but not so good at finding anything else.
Is any search system able to do more than one or two things well? Google does the ad thing. Lexis does the legal laundry list thing. Chemical Abstracts does the structure thing. Sure, these systems purport to provide more functions than a bucket of Swiss Army knives.
But the reality of search and information retrieval is that each system has a strength. Each system has gaps, blind spots, and stuff that just does not work as the users expect.
The write up identifies some of Facebook’s notable gaps; for example, dirty data. Don’t most Facebook users perceive content in Facebook is as accurate?
Net net: Facebook social search is a beta. What changes are coming? Wait and see.
Stephen E Arnold, February 16, 2013
Solr Unleashed Offered by LucidWorks
February 15, 2013
LucidWorks is a company offering commercial support, consulting, training, and value-added software to the open source Apache Lucene and Solr technologies. LucidWorks not only builds upon trusted open source technologies, it supports open source technology by employing committers. They also offer professional training on the open source components, even for those who are not interested in their LucidWorks Search or LucidWorks Big Data solutions. One such training opportunity is Solr Unleashed.
Read about upcoming classes:
“Having consulted with clients on Lucene and Solr for the better part of a decade, we’ve seen the same mistakes made over and over again: applications built on shaky foundations, stretched to the breaking point. In this two day class, learn from the experts about how to do it right and make sure your apps are rock solid, scalable, and produce relevant results. Also check the course outline.”
Register early for a discount on the two-day class. Opportunities are available stateside, as well as in Europe. Developers are the primary audience for the sessions, but system administrators can benefit as well. For more opportunities and to stay in the loop, contact the LucidWorks University team.
Emily Rae Aldridge, February 15, 2013
Sponsored by ArnoldIT.com, developer of Beyond Search
Sinequa France: Update 2013
February 14, 2013
My research team was winnowing our archive of information about European search vendors. Since Martin White’s article for eContent in 2011, a number of changes have swept through the search and content processing sector. Some changes were significant; for example, HP’s stunning acquisition of Autonomy. Others were more modest; for example, the steady progress of such companies as Sinequa and Spotter, among others.
The European technical grip on search is getting stronger. Google is the dominant player in Web search. But in enterprise content processing, some European firms are moving more rapidly than their North American or Pacific Rim counterparts.
The Sinequa tag cloud. See http://www.sinequa.com/en/page/solutions/category-1.aspx
One interesting example is Sinequa, based in Paris. The company, like other French technology firms, has a staff of capable engineers and managers. However, unlike some other companies, Sinequa has continued to establish a track record as a company innovating in technology and capturing some important accounts; for example, Siemens, the German industrial powerhouse.
Sinequa’s approach is to emphasize that enterprise search has moved to unified information access. A number of companies make similar claims. Sinequa has established that its technology can deliver the type of one-stop access to structured and unstructured content that almost every vendor claims to deliver. You can get a useful overview of the architecture of the Sinequa platform at http://www.sinequa.com/en/page/product/product.aspx.
A relatively recent addition to the Sinequa.com Web site are case analysis videos. I find case examples extremely useful. The presentation of this type of information in rich media format makes it easier for me to get a sense of the value of the solution a vendor delivers. I found the Mercer video particularly interesting. You can find these testimonials at http://www.sinequa.com/en/page/clients/clients-video.aspx.
The trajectory of European search, content processing, and analytics vendors is difficult to plot in today’s uncertain economic climate. Sinequa warrants a close look for organizations seeking an integrated approach to its content assets. For more information about Sinequa’s current activities, tap into the firm’s blog at http://blog.sinequa.com/
Stephen E Arnold, February 14, 2013
Sponsored by EMRxNow, the information service which tracks automated indexing of electronic medical records
-
- Subscribe to Beyond Search
Feature archive
News archive
-
