Mobile Search: Google in Line for a Swift Treatment
May 4, 2015
In the Jack Black “Gulliver’s Travels,” the protagonist found himself tied down by little people. I love Swift’s words for the small ones: Lilliputians and Blefuscusians.
I would recommend the 1735 edition, but I know that the one or two readers of this blog prefer to consume their history via videos.
The article “Search Start Ups See Opening to Challenge Google in Mobile” is the child of an earlier blog post “Start-Ups Try to Challenge Google, at Least on Mobile Search.” Both write ups drive a single point:
The GOOG is big, clueless, and vulnerable.
The write up did not include an illustration of the comedian Jack Black, but I provided that to make clear how the New York Times perceive Mr. Google.
I learned:
Europe’s competition regulator filed antitrust charges against Google on the belief that the company’s search business had become so powerful that it was pretty much impossible to compete with.
Okay, “pretty much.” The thread knitting together the examples in the article is the notion of “deep links to connect mobile applications the way websites are linked on the web.” Oh, so that’s what a deep link is. I find that connecting applications is one function, and deep linking is another. But, hey, let’s not disrupt the flow of the Swiftian analysis.
I highlighted the names of the Lilliputians mentioned in the article:
- Quixery
- Reley
- URX
- Vurb
Why are start ups bedeviling the data hungry giant? The write up reports:
Mobile also has several special challenges for an entrenched player like Google. With its constellation of apps and competing operating systems, mobile is a highly fragmented universe, making it harder for one company to index all of the most relevant information the way Google has indexed the web. Also, the answer to many of the most common — and lucrative — queries is neatly structured inside popular applications like Yelp, the local directory service, making it easier to create focused search products that are unlikely to sink Google but could give it “a thousand tiny leaks,” said Jeremy Kressmann, an analyst at eMarketer who covers the search business.
Yep, an expert who is an analyst in eMarketing.
But Google, according to the write up, may not be a dead mobile search goose yet. The write up says:
The company’s biggest bets have been a voice-searching tool, along with Google Now, an application that tries to predict what users are looking for by showing a stack of cards with timely information, using cues like coming events in the user’s emails or recent activities on mobile apps and the web. Like her start-up competitors, Ms. Chennapragada is still unsure exactly what people want. “Google Now is such an early effort,” she said. “We’re still trying to figure it out.”
Poor Google. Just not able to “figure it out.”
My thought is that vendors with mobile search technology may want to pursue a slightly different path than one that pesters the Google. SRCH2, another mobile start up founded by a Xoogler, is focusing on providing a service to larger outfits that need a mobile search solution.
My hunch is that the opportunity to suggest that Google is a vulnerable giant was more important to the write up than fiddling around with deep links and looking beyond the names of outfits with pipelines to eMarketers.
Stephen E Arnold, May 4, 2015
Mobile Yahoo Is Just Moving Right Along
May 4, 2015
What’s with the Yahoo? I know the company has been struggling with revenues. There have been executive shuffles. The mail service is wonky.
I read “How Marissa Mayer Mobilized Yahoo” in hopes of finding some answers to the mysteries of creating a new password for Yahoo email or configuring MyYahoo settings so actual content appears. I yearn for more information about Yahoo’s advanced search technology. Oh, I thirst for information.
I learned:
Yahoo believes there are signs that Mayer’s bet on mobile is paying off already. Some 575 million of its 1 billion users now access offerings such as the Yahoo app, Yahoo Mail, Yahoo Weather, Yahoo News Digest, and Flickr on mobile devices. Its rate of mobile growth outpaces the industry average. In 2014, the first year the company broke out mobile revenue, it reported grossing $1.2 billion in the category. “We had to build that—the people, the core competencies, the product base, the users, the traffic, and that revenue—from scratch,” Mayer says. “And we did it really quickly.”
Yep, mobile. Search and service stability?
The write up reports?
Employees in the company’s media division can complain that bureaucracy substitutes for a clear sense of direction. One ex-employee reports that people still at the company grumble of months-long waits for required approvals on projects. But at least in mobile, which is a top-priority focus, Yahoo does seem to be moving fast.
Yes, fast. Mobile.
The article states:
Almost every day, she says [Marissa Mayer] , she hears from people who are “surprised” by how good Yahoo’s mobile apps are. It’s time to stop being surprised. For the first time in years, Yahoo is positioned for some kind of future success, whether mild or bold. It’s not the company she inherited, and that’s a victory that no one—not even a cabal of angry investors—can take away from Marissa Mayer.
Well, maybe. What about search? What about that flashy information access technology? What about revenue growth? What about those acquisitions? What about juicy profits?
The answer is, it seems, “Mobile.”
Maybe.
Stephen E Arnold, May 4, 2015
The Guardian: Is Tech Envy Blooming?
May 4, 2015
The Guardian newspaper is one of the more tech forward newspapers which print dead tree editions. The publication has embraced open source. It features articles about technology and some of them are fact centric. One of the more interesting write ups or information constructs I have seen is “Take the Test: Could You Get a Job at Google?” I am one of the people who know the answer to this question.
I assumed the article would rehash one of the now hard to find Google Lab Aptitude Tests.
Here’s a sample page from an early version:
Here’s one of the questions:
Consider a function which, for a given whole number n, returns the number of ones required when writing out all numbers between 0 and n. For example, f(13)=6. Notice that f(1)=1. What is the next largest n such that f(n)=n? Don’t go running to WolframAlpha. Just write down the answer and no wonky faux flow charts.
Or you can work through the Guardian’s version of the test. Here’s a representative Guardian question:
Do you prefer dogs to cats?
Here’s a GLAT question.
What’s broken with Unix? How would you fix it?
I don’t have that experience with Googlers and Xooglers. Well, maybe a teeny, tiny bit. But I surmise the Guardian is poking fun at the GOOG. My hunch is that the write up is designed to amuse those at the Guardian. real Googlers and Xooglers are not likely to care. The write up reveals more about what the Guardian perceives about Google and the Guardian’s sense of humor.
Does anyone remember the commercial in which the key beat was, “It’s not nice to fool Mother Nature?” I wonder if it is okay to spook Mother Google. Of course it is. Would Google react to a person or organization with a Web site? Nah.
Stephen E Arnold, May 4, 2015
Indexing Rah Rah Rah!
May 4, 2015
Enterprise search is one of the most important features for enterprise content management systems and there is huge industry for designing and selling taxonomies. The key selling features for taxonomies are their diversity, accuracy, and quality. The categories within taxonomies make it easier for people to find their content, but Tech Target’s Search Content Management blog says there is room improvement in the post: “Search-Based Applications Need The Engine Of Taxonomy.”
Taxonomies are used for faceted search, allowing users to expand and limit their search results. Faceted search gives users a selection to change their results, including file type, key words, and more of the ever popular content categories. Users usually don’t access the categories, primarily they are used behind the scenes and aggregated the results appear on the dashboard.
Taxonomies, however, take their information from more than what the user provides:
“We are now able to assemble a holistic view of the customer based on information stored across a number of disparate solutions. Search-based applications can also include information about the customer that was inferred from public content sources that the enterprise does not own, such as news feeds, social media and stock prices.”
Whether you know it or not, taxonomies are vital to enterprise search. Companies that have difficulty finding their content need to consider creating a taxonomy plan or invest in purchasing category lists from a proven company.
Whitney Grace, May 4, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
EnterpriseJungle Launches SAP-Based Enterprise Search System
May 4, 2015
A new enterprise search system startup is leveraging the SAP HANA Cloud Platform, we learn from “EnterpriseJungle Tames Enterprise Search” at SAP’s News Center. The company states that their goal is to make collaboration easier and more effective with a feature they’re calling “deep people search.” Writer Susn Galer cites EnterpriseJungle Principal James Sinclair when she tells us:
“Using advanced algorithms to analyze data from internal and external sources, including SAP Jam, SuccessFactors, wikis, and LinkedIn, the applications help companies understand the make-up of its workforce and connect people quickly….
“Who Can Help Me is a pre-populated search tool allowing employees to find internal experts by skills, location, project requirements and other criteria which companies can also configure, if needed. The Enterprise Q&A tool lets employees enter any text into the search bar, and find experts internally or outside company walls. Most companies use the prepackaged EnterpriseJungle solutions as is for Human Resources (HR), recruitment, sales and other departments. However, Sinclair said companies can easily modify search queries to meet any organization’s unique needs.”
EnterpriseJungle users manage their company’s data through SAP’s Lumira dashboard. Galer shares Sinclair’s example of one company in Germany, which used EnterpriseJungle to match employees to appropriate new positions when it made a whopping 3,000 jobs obsolete. Though the software is now designed primarily for HR and data-management departments, Sinclair hopes the collaboration tool will permeate the entire enterprise.
Cynthia Murrell, May 4, 2015
Sponsored by ArnoldIT.com, publisher of the CyberOSINT monograph
Microsoft and Its Play in the Document Database Sector
May 3, 2015
The number of companies offering alternatives to Oracle’s traditional database management system continues to go up. The most notable anti-Oracle outfit may be Microsoft, or, the “new” post-Ballmer Microsoft.
A number of specialist companies have offered expensive and often quite complicated NoSQL content management systems. These have been positioned as publishing systems, analytic systems, business intelligence, and other types of unstructured information solutions.
I read “Microsoft Announces General Availability of Azure DocumentDB.” The key paragraph in the write up is a quote from a Microsoft expert which asserts:
We built DocumentDB in response to the increasing demands of mobile first, cloud first application development. NoSQL databases are becoming the tool of choice for many developers however running and managing these databases can be complicated and costly, especially at scale. DocumentDB is delivered as a fully managed database-as-a-service (DBaaS) with built in high availability, SQL query over indexed JSON and multi-document transaction processing.
Like most things Microsoft, this solution has been chugging along for many years. Microsoft has its own relational database to nurture. Leaving the market to the likes of MarkLogic and legions of open source developers and repackagers motivated the ageing Microsoft to keep on pecking away at the unstructured information problem.
Will Microsoft be successful?
The question is, “Sure, probably.” There are many Microsoft certified professionals, certified Gold resellers, and tag-along consultants who milk the Microsoft cow. The installed base combined with job security and consulting opportunities means that over time, Microsoft’s document database will have an impact.
What companies will be affected? There are several categories of firms which may suffer some immediate pain; for example:
- MarkLogic-type proprietary vendors will have to deal with the price pressure Microsoft brings to the market. Microsoft also supports a community familiar with the Byzantine methods of Microsoft software. The combination of familiar programming conventions and job security are likely to be a powerful mix.
- Vendors who drifted from Microsoft may have to fall in love again. Some specialist companies like Coveo and Smartlogic have shifted from Microsoft centric business to support for other systems and methods. If the document centric database sector catches on quickly, then these companies will have to rekindle their Microsoft affection. The challenge will be to fit into a Microsoft world in which more loyal developers have been faithful.
- Licensees reeling from the cost and complexity of XML centric document solutions may find that loyal licensees are kicking the tires of the Microsoft approach. Microsoft developers are easier to find and hire than masters of proprietary XML programming methods. JSON may become a good enough alternative to XML wonkiness.
What about search? I assume that an open source outfit like Elastic will encourage its fans to carry coals to the New Microsoft castle. Vendors of proprietary Microsoft add ins are likely to be revving up to offer snap in solutions to Microsoft’s migraine inducing search and retrieval function.
Net net: Microsoft is creating an opportunity as it inflicts pain on companies struggling in a tough economy. Worth monitoring this initiative. Will Oracle sit on the sidelines? How will proprietary content processing vendors respond? No answers yet.
Stephen E Arnold, May 3, 2015
Smart Software in the Cloud
May 3, 2015
If you have a need for analytics in the cloud, you may want to review the list of vendors in “10+ Cloud Machine Learning Platforms.” The list, assembled by a mid tier consulting firm, is eclectic. Amazon, Google, IBM Watson, and Microsoft are listed. Also included are services from Wise and BigML. The services are described with one major omission: The degree of technical expertise required. I assume that when one wants to “do” a machine learning project, the requisite statistical and programming skills are at hand. The challenge to the customer of these services may be significant.
Stephen E Arnold, May 3, 2015
Listen Up, Search Experts, Innovation Ended in 1870
May 2, 2015
I just completed a video in which I said, “Keyword search has not changed for 50 years.” I assume that one or two ahistorical 20 somethings will tell me that I need to get back in the rest home where I belong.
I read “Fundamental Innovation Peaked in 1870 and Why That’s a Good Thing,” which is a write up designed to attract clicks and generate furious online discussion and maybe a comic book. I think that 1870 was 145 years ago, but I could be wrong. Ahistorical allows many interesting things to occur.
The point of the write up is to bolster the assertion that “society has only become less innovative through the years.” I buy that, at least for the period of time I have allocated to write this pre Kentucky Derby blog post on a sparkling spring day with the temperature pegged at 72 degrees Fahrenheit or 22.22 degrees Celsius for those who are particular about conversions accurate to two decimal places. Celsius was defined in the mid 18th century, a fact bolstering the argument of the article under my microscope. Note that the microscope was invented in the late 16th century by two Dutch guys who charged a lot of money for corrective eye wear. I wonder if these clever souls thought about bolting a wireless computer and miniature video screen to their spectacles.
The write up reports that the math crazed lads at the Santa Fe Institute “discovered” innovation has flat lined. Well, someone needs to come up with a better Celsius. Maybe we can do what content processing vendors do and rename “Celsius” to “centigrade” and claim a breakthrough innovation?
Here’s the big idea:
“A new invention consists of technologies, either new or already in use, brought together in a way not previously seen,” the Santa Fe researchers, led by complex systems theorist Hyejin Youn, write. “The historical record on this process is extensive. For recent examples consider the incandescent light bulb, which involves the use of electricity, a heated filament, an inert gas and a glass bulb; the laser, which presupposes the ability to construct highly reflective optical cavities, creates light intensification mediums of sufficient purity and supplies light of specific wavelengths; or the polymerase chain reaction, which requires the abilities to finely control thermal cycling (which involves the use of computers) and isolate short DNA fragments (which in turn applies techniques from chemical engineering).”
Let’s assume the SFI folks are spot on. I would suggest that information access is an ideal example of a lack of innovation. Handwritten notes pinned to manuscripts did not work very well, but the “idea” was there: The notes told the lucky library user something about the scroll in the slot or the pile in some cases. That’s metadata.
Flash forward to the news releases I received last week about breakthrough content classification, metadata extraction, and predictive tagging.
SFI is correct. These notions are quite old. The point overlooked in the write up about the researchers’ insight is that pinned notes did not work very well. I recall learning that monks groused about careless users not pinning them back on the content object when finished.
Take heart. Most automated, super indexing systems are only about 80 percent accurate. That’s close enough for horse shoes and more evidence that digital information access methods are not going to get a real innovator very excited.
Taking a bit of this and a bit of that is what’s needed. Even with these cut and paste approach to invention, the enterprise search sector leaves me with the nostalgic feeling I experience after I leave a museum exhibit.
But what about the Google, IBM, and Microsoft patents? I assume SFI wizards would testify in the role of expert witnesses that the inventions were not original. I wonder how many patent attorneys are reworking their résumés in order to seek an alternative source of revenue?
Lots? Well, maybe not.
Stephen
Semantic Search and Dolphins
May 2, 2015
Do you remember Dolphin Search? I have some information in my Overflight archive. One Dolphin was a commercial search system. I made a note, “Based on an analysis of dolphin sounds.” I never followed up. I have another reference to an open source search system. Here’s a screenshot I snagged:
I located one of the gosling’s notes. The main point, “Flakey.”
Dolphins surfaced again (heh heh heh) in a write up with the tsunami of a title “How Dolphins Saved Semantic Search.”
This emergence of dolphins from the pool of information access is fascinating. Here’s the lead paragraph:
“We don’t talk about trust and identity much, it’ll be a good subject to discuss,”Teodora Petkova explained to me over an email. We were discussing my participation at the Sofia SEO Conference 2015 organized by Ognian Mladenov and my keynote opening speech on semantic search.
Yes, directly designed to hook me with a tasty morsel of bait or snare me in a mile long drag net.
I had to wait until the sixth paragraph to get to the dolphin and semantic search bobber.
I’d read about the Irrawaddy river dolphins while doing some research on human cooperation behavior. The Irrawaddy river is in Myanmar (former Burma) and the local fishermen had managed to find a way to communicate with the river dolphins. More than that the two of them had managed to create a shared language. The dolphins could, through their behavior, tell the fishermen just how big the catch they were bringing in was. The Fishermen would call out to the dolphins as necessary. This is not just cooperative behavior, it is a mutualistic relationship where both parties work together, sharing the workload so that the burden involved becomes less and the rewards for each, become greater. It is, in other words, contact between two intelligences. Search is similar.
Well, there’s that.
The write up drifts from “semantic” into a haze much like the mist that rises on some mornings from the south facing inlet on Guarujá. Where the write up drifts is the conflation of humanness and semantic search. The metaphor which drifts to mind is the plastic trash and detritus that disfigure beaches.
Well, there’s that.
Toss the chum of search engine optimization into the murky water and what emerges is the dolphin.
I think that this makes perfect sense to a person less wise than Hemmingway’s Old Man in the novella foisted on clueless sea farers in rural Illinois. The logic meshes with warnings like “red sky at morning, sailors take warning” or something similar.
Dolphins seem not to notice the weather. Dolphins do notice fish dangled in front of their noses at Sea World.
With semantic search playing the key part in this fish tale, the question arises, “What the heck does this have to do with information retrieval.”
Empty net. For sure.
Stephen E Arnold, May 2, 2015
Funnelback: Another Enterprise Search Solution Founder Offers Shoulds to Potential Licensees
May 1, 2015
Funnelback, as I have mentioned, has lost some of its marketing oomph. I think some staff shuffling took place. The company is now stepping up its effort to remain visible in a darned tough, crowded, and struggling market sector: Enterprise search.
I read “How Do You Solve a Problem Like Enterprise Search?” My answer is and has been, “One does not. One solves specific information access problems.” The wreckage of Convera, Delphes, Entopia, Fast Search, et al is evidence that enterprise search is a sticky wicket. The howls of pain on the LinkedIn forums and the odd collection of content in the Paper.Li round up about enterprise search make the challenges quite visible.
Read the article. Here are two points I found interesting.
According to the founder of Funnelback:
A holistic enterprise search solution should include:
- Bird’s-eye view metrics of all content, showing where it’s stored (e.g. web vs. enterprise vs. social media), how much exists in each repository, how old it is, missing metadata, poor quality titles, duplication, accessibility metrics, and the link graph. This provides information managers with a means to prioritize organizational investment in managing information, and thereby enhancing search effectiveness.
- Intelligent guidance on how to make content more visible/findable. Search engines generally attempt to hide the internals of their ranking systems and this makes it difficult for customers to learn how to make content more findable. An enterprise search engine should use its internal ranking knowledge to show content authors why pages rank the way they do and provide guidance on how to increase each page’s findability.
- The ability to surface and promote content based on user context with simple rules such as “User is in Department A”, “User is located in New Zealand”, “User is in the finance industry”, “User works for LexisNexis”. These rules can then be overlaid to form more sophisticated rules, without the need to create rules for every distinct possibility. Funnelback goes even further by allowing these rules to be applied to anonymous users by looking up their IP address in an internal database and inferring information based on the organization that owns the IP address.
These are darned interesting “shoulds.” The problem of access controls, contractual and regulatory constraints, and the human practice of creating silos of information are tough nuts to crack. “Shoulds” are easy. Delivering is tough, and Funnelback is neither more or less well equipped than open source or proprietary information retrieval solutions.
The second point illustrates the flawed logic that many champions of enterprise search as a grand solution make. Here’s the passage:
The first question every organization should ask is: Who are the stakeholders affecting the success of our organization and what information do they need to maximize our success?
At a more practical level, this includes questions like:
- What are the personas in our organization? (i.e. the archetypes that represent the different roles)
- What information do they need in order to maximize productivity and make better decisions?
- What are our customer personas?
- What information do they need in order to maximize engagement and have a positive customer experience?
Without asking these questions, organizations sometimes assume that searching everything with a single query (access controls permitting) is the answer. Sometimes it is the answer, but it can be a more complicated and costly exercise than necessary. For example, do users want to use an enterprise search tool to search their own email, or would they prefer to use the search on their mail client?
Sorry, Funnelback. Asking the questions is the first step. The work is to answer the questions and then use that information to tailor a solution that does not anger the users, lead to litigation, or just not work.
Today’s flagship enterprise search vendors seem to include Coveo, dtSearch, Elastic, Funnelback, and a handful of other firms with low profiles. The present crisis in information access has been created by the actions of previous industry leaders in enterprise search.
The fix is to focus on solving a problem for a specific group of users. Lawyers have specialized search tools. Chemists have specialized search tools. Regular employees have Google and whatever findabiliy solution is available within specific applications.
Want to get in a pickle? Sell a clueless senior executive a solution that solves the information access challenges for the entire organization. Didn’t work for STAIRS and won’t work for today’s systems.
The history of search is a painful one. There are options, but these are next generation systems, not yesterday’s systems wrapped with shoulds.
Stephen E Arnold, May 1, 2015
-
- Subscribe to Beyond Search
Feature archive
News archive
-
