China Trusts AI to Facilitate Human Communication
November 1, 2017
With the world’s largest population, one would think that the Chinese would not have any trouble finding someone to talk to. Apparently, China enjoys talking with robots, says the MIT Technology Review’s article, “Why 500 Million People In China Are Talking To This AI.” Going by the name iFlyTek, the AI app acts as on-demand translation service, but it does more than translate languages.
Over 500 million Chinese are using iFlyTek to manage their conversations with other people, including dictating texts, translating accents, transcribe, and generate automated messages. iFlyTek is programmed with all good tools related to communication: voice recognition, natural language processing, machine translation, data mining, and more. While the app has many applications in day to day life, the translation feature still has issues and intent is lost in translation.
The iFlyTek app is used in a variety of industries, especially healthcare and driving. Drivers issue it vocal commands sop their hands can remain on the wheel. Also, a hospital implemented ten female-looking robots to assist the overworked medical staff. The robots can answer questions and direct patients to the correct department. Doctors are also using iFlyTek to dictate a patient’s medical records. Dictation will become more important, especially since it offers people a hands-free way to get work done. There, of course, remains problems:
Although voice-based AI techniques are becoming more useful in different scenarios, one fundamental challenge remains: machines do not understand the answers they generate, says Xiaojun Wan, a professor at Peking University who does research in natural-language processing. The AI responds to voice queries by searching for a relevant answer in the vast amount of data it was fed, but it has no real understanding of what it says. In other words, the natural-language processing technology that powers today’s voice assistants is based on a set of rigid rules.
Vocal-based technology is becoming better, more accurate, and more reliable, but thee are still kinks in the system.
Whitney Grace, November 1, 2017
Natural Language Processing: Tomorrow and Yesterday
October 31, 2017
I read “Will Natural Language Processing Change Search as We Know It?” The write up is by a search specialist who, I believe, worked at Convera. The Search Technologies’ Web site asserts:
He was the architect and inventor of RetrievalWare, a ground-breaking natural-language based statistical text search engine which he started in 1989 and grew to $50 million in annual sales worldwide. RetrievalWare is now owned by Microsoft Corporation.
I think Fast Search acquired a portion of Convera. When Microsoft purchased Fast Search, the Convera technology was part of the deal. When Convera faded, one rumor I captured in 2007 was that some of the Convera technology was used by Ntent, formed as the result of a merger between Convera Corporation and Firstlight ERA. If accurate, the history of Convera is fascinating with Excalibur, ConQuest, and Allen & Co. in the mix.
In the “Will Natural Language Processing Change Search As We Know It” blog post, I noted these points:
- Intranets incorporating NLP, semantic search and AI can fuel chatbots as well as end-to-end question-answering systems that live on top of search. It is a truly semantic extension to the search box with far-reaching implications for all types of search.
- With NLP, enterprise knowledge contained in paper documentation can be encoded in a machine-readable format so the machine can read, process and understand it enough to formulate an intelligent response.
- it’s good to know about established tool sets and methodologies for developing and creating effective solutions for use cases like technical support. But like all development projects, take care to create the tools based on mimicking the responses of actual human domain experts. Otherwise, you may run into the proverbial development problem of “garbage in, garbage out” which has plagued many such expert system initiatives.
Mr. Nelson is painting a reasonable picture about the narrow use of widely touted technologies. In fact, the promise of NLP has been part of enterprise search marketing for decades.
What I found interesting was the Convera document called “Accurate Search: What a Concept, published by Convera in 2002. I noted this passage on page 4 of the document:
Concept Search capitalizes on the richness of language, with its multiple term meanings, and transforms it from a problem into an advantage. RetrievalWare performs natural language processing and search term expansion to paraphrase queries, enabling retrieval of documents that contain the specific concepts requested rather than just the words typed during the query while also taking advantage of its semantic richness to rank documents in results lists. RetrievalWare’s powerful pattern search abilities overcome common errors in both content and queries, resulting in greater recall and user satisfaction.
I find the shift from a broad solution to a more narrow solution interesting. In the span of 15 years, the technology of search seems to be struggling to deliver.
Perhaps consulting and engineering services are needed to make search “work”? Contrast search with mobile phone technology. Progress has been evident. For search, success narrows to improving “documentation” and “customer support.”
Has anyone tried to reach PayPal’s customer support or United Airlines’ customer support? Try it. United was at one time a “customer” of Convera’s. From my point of view, United Airlines’ customer service has remained about the same over the last decade or two.
Enterprise search, broad or narrow, remains a challenge for marketers and users in my opinion. NLP, I assume, has arrived after a long journey. For a free profile of Convera, check out this link.
Stephen E Arnold, October 31, 2017
Google Now Searches the Digital Shelves of a Library
October 27, 2017
Google can search the Internet in mere seconds and locate nearly every single bit of information on the planet. Except for ebooks you can rent for free from your local library…oh wait! According to the Digital Reader, that is now possible-read the news in the article, “Google Added Local Library eBook Listings To Search Results.”
Google has undertaken the challenge to add library catalogs to its search results, at least the digital collection holdings. Can you imagine visiting Google to search your library’s catalog? It will now be a thing, so go ahead and search your local library holdings through Google.
Google, however, cannot access all of a public library’s holdings, the physical items are left out. There are also some bugs in the new feature, such as an item’s availability status:
I did run into some inconsistency with this new feature. While some books would bring up the Borrow ebook card without any problem, other titles just wouldn’t trigger it. While Cixin Liu’s “Three-Body Problem” had the extra field in results, searches for N. K. Jemisin’s recent Hugo Award winner “The Obelisk Gate” didn’t trigger the same behavior.
I first thought it might have to do with availability, but the card appeared for titles with universally long waiting lists, too. Behavior was the same for search on both desktop and mobile, and I’m at a loss to explain why.
It is not perfect, but this is an excellent way for Google to connect people not only with a library’s digital collection but the physical collection as well. Searching a library’s catalog on their website, however, still remains the best way to locate the material.
Whitney Grace, October 27, 2017
Equifax Hack Has Led to Oracle Toughening Up
October 19, 2017
According to a timely piece in SearchOracle, its parent company has muscled up in response to its recent troubles, according to the article, “Machine Learning and Analytics Among Key Oracle Security Moves.”
This comes on the heels of the infamous Equifax hack, which was made vulnerable due to a weakness in Apache Struts. To their credit, Oracle has owned up to the problem and made it public that they are not going to wilt in the face of criticism. In fact, they are doubling down:
Oracle’s effort to help IT teams reprioritize their defenses, he said, takes the form of a new unified model for organizing data, rolled out as part of an updated Oracle Management Cloud suite. Advanced machine learning and analytics will enable automated remediation of flaws like Struts…
The story continues:
(Oracle’s) approach to machine learning is uniquely its own, in the sense that it is being delivered as a core enhancement to existing offerings, and not as a stand-alone technology that is personalized by a mascot or nickname — a la Einstein from Salesforce or Watson from IBM.
We like that Oracle isn’t trying to throw the baby out with the bathwater, here. We agree, there are a lot of things to like and overhauling would not be the solution. Via analytical improvements, we suspect that Oracle will recover from the Equifax snafu and be stronger for it. They certainly sound like their focus is on that.
Patrick Roland, October 19, 2017
Google Eyes AI in China
October 19, 2017
Google wants to elbow its way into Chinese markets. As the most populous country in the world, China is more or less an untouched by Google and it is a big, potential market for the technology company. The South China Morning Post shares that Google is trying to get another foothold in the country: “Google Looking To Build Its Own Team Of Artificial Intelligence Specialists In China.” Google wants to build an artificial intelligence team in China as the country plans to become the world leader in AI by 2030. It is not unthinkable that China could obtain this title as the country is making advances in other technology fields.
Google wants a piece of the Chinese profit pie and their job postings Web site lists several opportunities in the Middle Kingdom. Google has been posting AI job postings based in China since May. China is not only a populous country, it is also one of the most powerful countries in the world and is stretching its wings in order to compete with more technological heavy countries such as the United States, Canada, Western Europe countries, South Korea, and Japan.
With new technological innovations comes the potential for big profits and China is ready to invest and earn. China’s artificial intelligence market is predicted to grow 50% each year, which is much higher than the 20% the world market is expected to grow. Google has plans to be part of that picture:
The timing of Google’s recruitment drive for artificial intelligence specialists on the mainland coincided with its creation in May of a business division, called Google.ai, dedicated solely to that field. At Google’s annual developer conference held in San Francisco that same month, company executives announced a range of artificial intelligence initiatives. These included adopting machine learning to enhance Google Search and enabling Google Maps to automatically detect street signs.
Google has pressured China in the past to change its policies or the company would not conduct business with it. Google, however, has caved to China’s demands and has changed itself. It does not come as a surprise because Google is profit and research driven and China is an untapped market for it. Emphasis on the profit portion.
Whitney Grace, October 19, 2017
Artificial Intelligence Will Make Humans Smarter at Work
October 17, 2017
Relatively no industry has been untouched by the past decade’s advances in artificial intelligence. We could go on and make a laundry list of which businesses in particular, but we have a hunch you are very close to one right now. According to a recent Enterprise CIO article, “From Buzzword to Boardroom: What’s Next for Machine Learning?” human intelligence is becoming obsolete in certain fields.
As demonstrated in previous experiments, no human brain is able to process as much data at comparable speed and accuracy as machine-learning systems can and as a result, deliver a sound, data-based result within nanoseconds.
While that should make you sit up and take notice, the article is not as apocalyptic as that quote might lead you to believe. In fact, there is a silver lining in all this AI. We humans will just have to work hard to get there. The story continues:
It must also leave room for creativity and innovation. Insights and suggestions gained with the aid of artificial intelligence should stimulate, not limit. Ultimately, real creativity and genuine lateral thinking still comes from humans.
We have to agree with this optimistic line of thinking. These machines are not exactly stealing our jobs, but forcing humans to reevaluate their roles. If you can properly combine AI, big data, and search for your role, chances are an employee, like yourself, will become invaluable instead of obsolete.
Patrick Roland, October 17, 2017
Everyone and Their Dog Is a Search Expert
October 13, 2017
Young people get frustrated when they help older people with technology. There are considerable sighs, rolling eyes, and the situation often ends in yelling. One frustration young people are forced to deal with is teaching an older person how to use a search engine. Trying to explain how to enter information into the text box, the meaning of keywords, and how to tell the difference between results is not easy. However, search engines like Google, Bing, and Yandex try to make the search process as easy as possible so everyone can become a search expert.
Learning how to search is not the only thing people have trouble learning. Tech Viral wrote about the top “how to” searches in the article, “Here Are The Top 100 ‘How To’ Searches That People Want To Know.” Xaquin GV researched how people use Google as the answer all “how to” tool and discovered the most popular searches. Among the top “how to “searches are how to make money, how to tie a tie, how to draw, how to kiss, how to lose weight, how to make pancakes, and how to get pregnant.
The essay also examines the top 100 ‘How to’ searches conducted worldwide, and the results are very illustrative. Xaquin divided those searches into categories, with visual representations of how popular each of them is.
The search results mostly revolve around activities that are adult responsibilities along with a few surprises that concern current trends. Everyone can become an expert at any activity with a few simple keystrokes and tutorial guides. YouTube makes “how to” guides more helpful and even more dangerous when people try to copy the experts at parkour, skateboarding, and daredevil activities that should never be tried at home kids.
Whitney Grace, October 13, 2017
Google Search Data Utilized for Financial Analysis
October 6, 2017
Here is a short honk to point out an interesting new use for search data. A financial analyst is relying on it to make a key prediction, CNBC reports in, “Analyst Predicts Great Amazon Sales Results Because of What He Sees in Google Search Data.” Reporter Tae Kim writes:
Piper Jaffray’s Michael Olson reaffirmed his overweight rating for Amazon, citing the company’s web search analysis which pointed to robust June quarter sales growth for the e-commerce giant. …
Olson said the firm’s web analysis revealed search interest for Amazon-related words grew 24 percent year over year in the June quarter versus 23 percent growth in the March quarter. He cited how Piper’s search analysis had a 95 percent correlation with Amazon’s retail sales unit growth in the previous 37 quarters.
Such interest may be spurred by Amazon’s purchase of Whole Foods, and by the company’s strong growth in this year’s second quarter. The innovative analyst’s company, Piper Jaffray, has been in business since 1895. It is nice to see a venerable firm embrace a fresh idea, but will Olson’s prediction prove correct?
Cynthia Murrell, October 6, 2017
The Future of Visual and Voice Search
October 4, 2017
From the perspective of the digital marketers they are, GeoMarketing ponders, “How Will Visual and Voice Search Evolve?” Writer David Kaplan consulted Bing Ads’ Purna Virji on what to expect going forward. For example, though companies are not yet doing much to monetize visual search, Virji says that could change as AIs continue to improve their image-recognition abilities. She also emphasizes the potential of visual search for product discovery—If, for example, someone can locate and buy a pair of shoes just by snapping a picture of a stranger’s feet, sales should benefit handsomely. Virji had this to say about traditional, voice, and image search functionalities working together:
A prediction that Andrew Ng had made when he was still with Baidu was that that ‘by 2020, 50 percent of all search will be image or voice.’ Typing will likely never go away. But now, we have more options. Just like mobile didn’t kill the desktop, apps didn’t kill the browser, the mix of visual, voice, and text will combine in ways that are natural extensions of user behavior. We’ll use those tools depending on the specific need and situation at the moment. For example, you could ‘show’ Cortana a picture of a dress in a magazine via your phone camera and say ‘Hey Cortana, I’d love to buy a dress like this,’ and she can go find where to buy it online. In this way, you used voice and images to find what you were looking for.
The interview also touches on the impact of visual search on local marketing and how its growing use in social media offers data analysts a wealth of targeted-advertising potential.
Cynthia Murrell, October 4, 2017
Search and Privacy: A Quick Update
October 3, 2017
In my files, I had a copy of “Duck Duck Go: Illusion of Privacy.” This document comments on the hurdles a public Web search system must jump over in order to deliver privacy. You can find the write up at this link. If you want to test some privacy-oriented search systems, there are some DuckDuckGo.com alternatives. I am not endorsing these outfits; I am passing along some links because within the last couple of years I learned that privacy is part of the marketing for these systems: [a] Ixquick which is now Startpage at www.startpage.com. This is a metasearch engine which means that the user’s query is passed (in theory anonymously to Bing, Google, Yandex, et al). [b] Unbubble.com (Note that this European service asserts “strong privacy.” The link is www.unbubble.eu [c] Gibiru service (www.gibiru.com) emphasizes anonymous search. Gibiru provides a link to the Firefox Anonymox plug in. But the most recent version of Firefox has been tricky for us, however. My personal view on search anonymization is that when I research my books about cyberosint, the Dark Web, and eDiscovery for cyber intelligence, I assume that I have a number of individuals thrilled with the sites we uncover, write up, and describe in our lectures and webinars. In short, I avoid trying to be “tricky” because I can explain the thousands of queries we run about many exciting topics. See www.xenky.com/darkwebnotebook for a sampler.
Stephen E Arnold, October 3, 2017
-
- Subscribe to Beyond Search
Feature archive
News archive
-
