+1-312-477-7300
eContext

Storytelling and Crowd-Sourced Taxonomy

by

Last week, one of our favorite things (the library) teamed up with one of our favorite radio shows (The Moth) using one of our favorite technologies (speech-to-text) to grow a taxonomy of storytelling. We’re elated!

If you’re a fan of The Moth, you know it’s “storytelling, live, without notes” — a live event, radio show,  podcast, and more that features real people telling stories on a stage. Stories range from raw and tragic to deeply meaningful and wildly funny.

Since launching in 1997, more than 20,000 stories have been told from Moth stages across the U.S., and now, throughout the world. In an effort to better curate their collection of stories and make them available to people with disabilities such as hearing impairment, The Moth set out to find a partner who could help create transcripts.

This led The Moth into a collaboration with The New York Public Library and a speech-to-text company called PopUp Archive. With a grant from the Knight Foundation, the three organizations collaborated on a very interesting strategy for creating and archiving transcripts of The Moth.

Turning Speech into Natural-Language Text

The problem with speech-to-text programs is, well, sometimes they don’t hear very well. In addition to basic language barriers such as accents, dialects, and fluctuations in speech, which occur all the time, Moth recordings typically take place in a bar where there is a lot of ambient noise, varying levels of audio quality, and other factors that can make a speaker’s words difficult for a machine to accurately recognize. This means that transcripts produced by PopUp Archive would need a good edit, and we’re guessing the Knight Foundation’s grant wasn’t going to cover editing expenses for 20,000 files.

The New York Public Library (NYPL) had a solution: Crowdsourcing. Turns out NYPL Labs had developed an interactive transcript editor called Together We Listen, which allows anyone to log on, listen to Moth stories, and assist with editing the transcripts as they go. Volunteers smooth out the transcripts by looking for misspelled words, or misplaced synonyms, putting words and phrases into the context that the speaker intended as they go. Volunteers are also asked to add tags to each file to make it easier to catalogue and categorize the collection.

A Taxonomist’s Dream Come True

We probably don’t have to tell you how much we love this idea. As people who sit in an office all day categorizing words and phrases, we’re thrilled to learn that people are out there volunteering their time to curate stories. The categorization element exemplifies the quiet work that taxonomists do every day in order to curate and maintain information.

The project also illustrates the value of classification on a very practical level. By transcribing Moth stories into text, the organization is making the stories more accessible to all of us, particularly those who may not have been able to enjoy Moth stories before they were available in print. And, because these stories represent such a wide array of topics and audiences, we’ll now be able to reach into the archive and pull out the categories or topics that we want to learn about. Additionally, The Moth will be able to use the catalogue in a number of ways to enrich its offering. For example, they can dig deep and find collections of stories to publish, such as stories from cancer survivors, or humorous stories that will make us laugh. They could pull together various collections to package and sell as educational tools, or deliver favorite celebrity stories online as sponsored content.

We’ve curated more than 450,000 terms, and we’re here to tell you, it can sometimes be a lonely business. Knowing that there’s a growing community out there that loves to categorize, curate, and organize information into a hierarchy makes us smile.

In June, the NYPL is hosting a hackathon and storytelling celebration. We’ll provide more details when we hear about it. In the meantime, be part of the solution by helping to curate these stories!

by

Taxonomy is the heart of eContext. Whether you’re using our technology to expedite machine learning projects, segment audiences by interest, or develop conversational interfaces, a curated subject matter hierarchy is what makes our semantic classification both intuitive and flexible.

Of course, many of the businesses we work with come to us with their own existing topic structures—some homegrown, some open-source. If you need an easy example, just think of any shopping website and the navigation panel you’re sure to find on the left side or top of the screen.

So the obvious question is: if companies already have their own taxonomies, how can they utilize ours? Some prospective customers have expressed concern that, in order to incorporate our text classification capabilities, they’d have to ignore their native structures in favor our ours.

Well, not to worry: there’s no need to throw the baby out with the bathwater. One of the services available to our clients is a taxonomy overlay, which means we map our categories against yours, so subsequent classification efforts are geared towards your unique structure.

How does it work?

Step One: Classify

We pass your existing structure, including any relevant metatags or descriptions, through eContext’s classification engine. The output shows which of your organizational categories correspond directly with our own for the simplest of 1-to-1 conversions. We’ll then review the results and make sure there aren’t any errors.  If we need to build any new categories into the eContext taxonomy, that will happen at this stage as well (though it’s rare that you’ll have a subject in your structure that we don’t have in ours.)

Step Two: Extend

 

We’ve been perfecting our category structure since 2008, working throughout that time to improve in both breadth and depth. Therefore, we typically find that the eContext taxonomy is much more comprehensive than other organizations’, covering a wider variety of subjects and segmenting topics with deeper tiers of subcategories.

Thus, clients receiving an overlay have the option to augment their existing structures with topics or segmentations they may be lacking. Suppose you’re a sporting goods retailer, and your organizational structure goes as deep as, say, Sporting Goods >> Winter Sporting Goods >> Skiing Equipment. eContext could map these categories against our taxonomy as described in step one, but we could also extend this branch to include several subtypes of skiing equipment, such as Ski Poles, Ski Boots, or even Ski Wax.

Step Three: Benefit

With your structure reconciled with ours–and extended where necessary–the vocabulary rules that govern our text classification can automatically be applied without any changes required on your end. Now, any new data you classify via eContext will fit your current setup, which is great if you’re taking in more data than you’re able to sort manually. Besides streamlining your existing operations, taxonomy overlay has the added benefit of future-proofing your data, minimizing complications that arise from incorporating new software or merging with another organization.

This is a pretty high-level rundown; there may be unique considerations (and advantages) in performing a taxonomy overlay for your business. We’d love to discuss with you, so feel free to contact us with any questions.

by

When you work on a product or service that’s primarily aimed at businesses, it’s easy to get accustomed to the well-traveled language of B2B.

eContext’s text classification, we’re used to saying, delivers insights. It empowers organizations to expedite their initiatives and adds value to big data by mapping assets to a hierarchical structure.

To be clear, all of these things are true. But a glut of buzzword-laden value statements–and let’s be honest, some have more meaning than others–can sometimes de-emphasize what makes an offering truly necessary or unique.

So this week, we decided to refresh our appreciation for eContext by looking at it from a different perspective. We asked a few staff members around the office: what if eContext were actually a consumer product? What if, instead of enriching data for large businesses, you were able to take advantage of topic classification in your day to day life?

The responses were great–a few people envisioned eContext as a kind of mobile app, while others wished for a browser extension or plugin to automatically classify content.

All-in-all, it was a fun exercise that reminded us how helpful it can be to have access to topic classification on a day to day basis.

Check out some of our favorite responses below.

 

Managing social media with better text classification

 

“I’m sick of apps, but what I’m even more sick of is social media.  I’ve deleted or abandoned most of my accounts, aside from Facebook, and that’s mostly to keep in touch with out-of-town friends and relatives, and to find out about local events.  I know there are ways to block a certain person’s content, but sometimes I don’t mind seeing cute pics of their new puppy, I just don’t wish to read their 82nd political post of the day.  Anecdotally, I’ve heard others complain that social media makes them “feel like crap” because it makes people’s lives look so much better than theirs.  For example, if someone just went through a break-up, they might not want to read about their college friend’s awesome spouse.  It would be useful to have an app that could filter out posts with certain keywords or topics, rather than filtering out people completely, and the app could include posts from various social networking sites, so there would be no need to log in to five of them.” — Allison, Taxonomist

 

Keeping recommendations in order

 

“I wish I had a better way to keep track of all the recommendations my friends and family pass along. TV shows, articles, book suggestions, that kind of thing. Usually, I try to make notes for myself on my phone or scrap paper, but it would be cool to have an app that could sort all that content by subject matter and keep it organized for when I have the time. Say I was between books and felt like reading a western, I could go into my app and see if my friends had given me any corresponding recommendations. Basically, I wish I had an eContext app to keep me more organized and account for my forgetfulness.” — Patrick, Administrative

 

Better networking  

 

“I’m always looking for ways to connect. I have a small business and I rely on networking for most of the work that I get. I’d love to be able to pull my Twitter feed and find out what people are saying about manufacturing or technology, two topics I work with frequently, so I could connect with them on a more effective level. Right now, it’s difficult to do this efficiently on Twitter, where you can’t get much deeper than a hashtag. I’d be interested to find ways to pull all of my friends who just happen to be interested in a topic like manufacturing and think of new ways to connect with them.

It’s also interesting to me to know the tangential things that people are interested in and finding unlikely associations. It’s simple enough to think of the obvious – a person who works at a manufacturing firm is also interested in manufacturing technology. I think it’s more fun to find unlikely associations, or things you wouldn’t necessarily think of, and eContext’s text classificaiton is a great way to visualize data and look for patterns that emerge. I would use this kind of information to find off-the-beaten path ways to reach out to potential new customers. If I can find marketing managers at manufacturing firms I suppose I could reach out to them with a message about marketing and manufacturing. But if I can find all the marketing managers at manufacturing firms who like the Cleveland Indians, I have a much better chance at getting their attention if I can wedge my Cleveland connection (and love of the underdog team) into my introduction. It’s very effective, but it takes time to do that kind of research. eContext could make it easier.

Go Tribe.” — Miriam, Marketing

by

Millenials are showing greater interest in voting, and have a shot at changing the political landscape. Heightened interest means greater online traffic, and more opportunity for publishers to court and capture new readers and fans.
The Pew report shows, on the one hand, how much division there is in the types of sources that Americans turn to for news and information, but it also reveals how important social media, news websites, and apps are to the coveted younger audience. In the graph below, almost 50 percent of respondents aged 18 – 29 report that the most helpful news they get about the election comes from social media or news websites and apps.

PJ_2016.02.04_election-news_0-02

Capturing and keeping Millennials

Publishers are under a great deal of pressure to keep readers engaged and moving through their site, but often they’re using clunky, third-party tools or widgets to what a reader should read next, or worse, they’re asking blog contributors or lower-level data entry employees to manually select categories and keywords for this purpose.

You’ve got a very short time to entice a reader to stay on your site, and the key to keeping them there is delivering relevant content options. If a Millennial winds up on your site trying to find out more about Hillary Clinton’s economic platform, and the related links on your site have nothing at all to do with Hillary, or economic policy, or even politics, you’re not going to get her to stay.

Consider this set of recommendations from the bottom of a Forbes.com post titled Hillary Clinton Rolls out her Economic Platform which, as of February 16th, promoted three recommended articles on the Forbes site, one on Oracle’s cloud service, a photo essay of the richest person in each state, and another piece on how to exceed customer expectations. Nothing on economics or politics.

Forbes’ own content recommendations don’t quite strike the right note, either. The site’s “see also” list is as follows, with the most relevant topic at the bottom of a list of hit and miss matches:

    • Best Investment Plans
    • Economic Forecasts
    • Oil Gas INvestments
    • Most Profitable Business
    • Future Gold Price
    • Money Management Tips
    • Senator Hillary Clinton
  • Hillary Clinton Biography

At the bottom of the page is a series of brightly colored, but even less relevant ads delivered to the Millennial who searched for “Hillary Clinton’s Platform” – do you think any of these items are going to keep our politically-motivated visitor clicking?

    • Ever Googled yourself? Do a Deep Search instead
    • Anxiety? Hemp Extract, Now Legal in IL
    • Forget Pensions! 11 Rules for Getting Rich After 50
    • Skip the Mattress Store – 3 Ways You’re Getting Screwed Buying Retail
    • 2 Dudes on Shark Tank Reveal A Trillion Dollar Idea
    • 10 Online Dating Sites that Really Work
    • A Brilliant Way to Pay off Your Mortgage (It’s Genius!)
  • Men: 5 foods That Kill Testosterone and Cause Belly Fat

Seems like the word “economic” may have triggered a shallow response and delivered a series of ads that are way out of the scope of what the searcher was looking for.

Building relevance into your site

Forbes is missing opportunities for added page views. Even if we have no additional information, we know that our Millennial is hoping to learn about Hillary’s platform, and while there is plenty of information available on the Forbes site about the election and Clinton’s views, there’s very little for the visitor to discover from this page, and only one suggested topic and not a single ad has to do with the reader’s original query. Why should our impatient twenty-something dig through Forbes when she can just hit “back”, search Google or Bing, and get the info she’s looking for on someone else’s web property?

It’s another example where text classification can be a huge help. Using eContext, Forbes could classify both user searches as well as site content to a single hierarchy of topics — topics such as “Hillary Clinton”, “Public Policy”, or “Economic Sciences”. By connecting the topics a user is searching for with the content in its inventory, Forbes can deliver truly relevant articles and more intelligently place sponsored content. Each page the visitor reads that delivers more relevant content helps the reader build more confidence in Forbes as a destination, and makes her more likely to click on sponsored content and/or return to the Forbes site in the future.

eContext’s taxonomy facilitates interaction on your site by helping you deliver relevant topics. There are more than 450,000 categories, reaching down 21 tiers to help you pull out the topics that are most relevant to your search now.

A recent Time article published by Jared Kleinert and Stacey Ferreira, How 2016 Candidates Can Win Over Millennials, strongly encourages candidates to use ad targeting platforms that will appeal to Millennials. Publishers who can deliver relevance in content and ad delivery will please readers, but their sites will also be a more appealing place for campaign ad spend. This election season, digital ad spending will represent a 5,000 percent increase over the 2008 election and is estimated to reach $1 billion, according to Wired, so making efforts toward more relevant results has the potential to deliver big numbers, particularly if you cater to Millennials.

This election season promises to deliver a great deal of drama, debate, and interest — think of what you can gain by using eContext to help your site deliver better options to readers.

by

Your customers are trying

As a retailer, it can be a struggle sometimes to bridge the gap between the items you sell and the customers who might buy them. So often you’re stuck in the middle with a product that a manufacturer has shipped to you along with the appropriate SKU and possibly a photo and a short description, and you have to figure out how to market that product so that your customers will find it on their first try when they get to your website.

One of the problems you have is that your customers aren’t looking up a formal SKU number — they’re going to your search box with the vocabulary they’re accustomed to using. If a customer arrives on your site and enters the keyword “strappy Choos”, will your search engine–trained to look for certain words from product descriptions–return any helpful results?. Sure, you sell plenty of strappy shoes, and you may even carry the Jimmy Choo brand she’s looking for,  but if your search results can’t make the connection, you’ve just given your customer a reason to leave your site.

 

Matching meaning instead of words

If your customer was visiting a retail location, a sales associate would take her right to the Jimmy Choo section, or show her a grouping of equally fun party shoes. But how can you recreate this experience online?

eContext has a solution that’s powered by human thinking. We can take your customer query and map it to our deep taxonomy to help you deliver better search results. We’ve had humans curating our taxonomy so that the terms are accurate and reflect real English language speech. Our classification goes up to 21 tiers deep and features 450,000 categories.  

When a shopper arrives at your site with a search term that’s puzzling to your search function, like “strappy choos”, pass that search term on to us. eContext looks at the search term in real time and returns category suggestions that your site logic can understand. Now you can deliver relevant options for your customer to consider, including “Jimmy Choo Footwear”,  “Heels”, and “Fashion Sandals”. With intelligent classification, what might have been a roadblock that sends your customer to competitor’s site is, instead, an intuitive shopping experience and a new opportunity for a sale.

In addition to real-time search, you can use eContext technology to classify product descriptions and consumer reviews. By applying the same topic classification to all of your interconnected product and consumer data, you’ll be well-equipped to match customer requests with relevant results.

With eContext, you increase your opportunities to sell to customers. How many times do you think you’ve lost a customer due to poor search? People will use search terms that make sense to them and it’s up to you to get to the heart of what they’re asking for and deliver.

 

Additional options with eContext

eContext helps you build more flexibility into your site by connecting you with a deep taxonomy that can roll up or roll down as you need it to. For example, maybe you carry a wide variety of Jimmy Choos, but want to promote the top-selling subcategory of “Jimmy Choo Ivette”. On the other hand, if you don’t carry the Jimmy Choo brand at all, you can move up a few categories and offer results for the more general topic of “Heels”.

eContext also returns scores, confidence measures for fit of a particular document to a category. These scores can be used for ranking. So, if the consumer is searching for something in the sandals category, and the retailer has many product descriptions related to that category, the scores can be used to select the most relevant product descriptions.

Data-savvy retailers can layer in other types of analytics as well, such as integrating relevance scores with click-through or conversion rates to identify the best choices for use in search results.

If you can build more relevance into your site, your end-user will stay longer, and you’ve got a much better chance to covert. If the shoe fits…

by

Las Vegas casinos will soon be full of the world’s leading data crunchers.

 

This week, the city will play host to the Big Data Innovation Summit, a two-day forum on advanced theories and applications for big data. . With that much mathematical brainpower floating around, there’s bound to be a few card counters working the tables during the conference off-hours.

 

If you skim through the Summit’s speakers and topics list, you’ll notice a few recurring themes: big data (obviously), machine learning, and IoT. But there’s another focusthat emerges over and over again, one that might seem out of place if it weren’t so ubiquitous: the human element.

 

Technically speaking

Beena Ammanath, executive leader of data and analytics for General Electric will be there to argue on behalf of industrial applications for big data, “It’s in the industrial area that big data is going to have the biggest impact, transforming economies, saving lives, reducing power consumption, changing the way we live,” notes her presentation overview – these are highly aspirational and human goals for a corporation to have, and the burden falls on the analytics team to achieve them.

 

Boeing is also looking for the human element, and will present on big data technologies they’ve developed and the important role of humans throughout the process.

 

Microsoft’s James McCaffrey, a senior research scientist, is presenting on neural networks. He is promising to explain how neural networks operate without using “Greek letters or annoying jargon.” It’s a telling hook: artificial intelligence used to be a purely esoteric concept, but in 2016, it’s becoming something that increasingly affects — and is informed by — the actions of ordinary people.  

 

Speaking of ordinary people, Arjit Sengupta, CEO of BeyondCore is talking about how his company interprets big data in a way that non-scientists can quickly understand. He’ll show how analytics is moving to a place where “users can easily overlay human intuition on top of automated analysis.”

 

Man vs. machine

We hear and read a lot about the threat of artificial intelligence to humanity — a matter of science fiction that geeks just love to talk about — but pop culture tends to minimize the truth that the human element will always play a critical role in the development and application of new technologies.

 

Everyday communication is a great example. Attendees at the Big Data Innovation Summit will probably talk about the need for better natural language processing and solid text analytics strategies, because real-world communication is such a complex, idiosyncratic system of signals.This is where human input plays a role, because in order for Big Data to deliver, machines have to be able to learn from us about what we mean when we speak and write. Machines need to know how to interpret informal, spoken language, slang-filled social media text, when we’re joking, when we’re quoting Caddyshack… it’s a complicated job. .

 

Our technology supports big data by bridging the gap between formal business English and everything else, helping to bring context to the new kind of data that businesses rely on to make better decisions. As the world’s largest text classification engine, we support any big data effort by bringing structure to the seemingly unstructurable.

 

For the last ten years, we’ve had humans curating a massive taxonomy than runs 21 tiers deep and contains almost half a million total categories. We can take data from any source and run it through our universal topic hierarchy in real time and deliver structured data for your applications. It’s simple and elegant, and it creates an opportunity to take a very deep dive into the human meaning of the data that you’re looking at.

 

As the number crunchers assemble in one of the most data-driven vacation and conference spots on the planet, we’ll be following  the conversation about the importance of human language to Big Data. Enjoy the show!

by

At the National Retail Federation’s Big Show this week, retailers are already looking at the coming holiday shopping season. The numbers are out for the 2015 retail holiday season and spending was strong. According to MasterCard Advisors SpendingPulse, U.S. retail sales grew by 7.9 percent year-over-year for the period of time between Black Friday and Christmas Eve.

The news was even better for online retailers. eCommerce grew 20 percent compared with last year, according to MasterCard.

With the rush of the holidays comes the chance to make or break annual sales goals in a matter of weeks. It’s an amazing opportunity for retailers, but can also be a source of terrible anxiety. According to the National Retail Federation (NRF), holiday sales as a percentage of annual sales for the average retailer is almost 20 percent. With so much at risk, retailers need to pull from every available resource to sell during the holiday season.

Zeroing in on what customers want

Retailers put a great deal of effort into preparing for the holiday season, and as online usage increases — MasterCard says 70 percent of U.S. consumers reported doing more research online than before — it’s more important than ever to have good digital marketing practices in place, paired with an amazing search experience that keeps shoppers engaged.

Retailers have honed their online marketing skills and have become adroit at attracting visitors to their sites through online advertising, advanced email, and social campaigns, but keeping visitors there from the first click to the the final checkout can be difficult.

Even very sophisticated sites often have trouble translating product codes and descriptions into search results that meet exactly the intent of the shopper. Often, this is due to a product-centric classification system that is helpful to back-end developers, but delivers results to shoppers that make no sense. Have you entered a seemingly straightforward query into a retailer’s search box, only to be treated to a slew of obscure parts or irrelevant products? When shoppers get bad results on their first search query, and then again on their second, you can bet they’re going to shop somewhere else.

Delivering context to search results

We’ve developed a classification system that focuses on matching the meaning of a word, not just the word itself. So, if a customer arrives at your site with high hopes of getting a VitaMix blender, which perhaps you don’t carry, eContext is smart enough to look into its 450,000 categories and 21 tiers to understand that what your customer is looking for belongs in this deep category:

Home & Garden >>Home Appliances >> Home Appliance Products >> Kitchen Appliances >> Blenders, Juicers, Smoothie Makers & Frozen Drink Machines >>Blenders

If you don’t offer VitaMix, or if the price tag for a VitaMix is a bit high for your shopper (those professional-grade blenders are a bit pricey), eContext can help you deliver suggested results that are similar by drilling up or down a level. Now, you can deliver results that might better fit your shopper’s price range or intended use.

With this kind of advanced classification system, you can learn a lot from your customers’ searches. Layer on your own analytics programs to discover visitor behavior and develop new ways to present products to visitors, and use that same intelligence to deliver better ads to get visitors there in the first place. eContext is a universal tool, so it can be used to classify your product inventory or search data.

When does planning for next holiday season begin? Right now. Smart retailers are pulling all the data they can from this past season and analyzing it so they can start planning for next year. You should too.

Ready to learn more?

Schedule a demo

Chicago

116 West Hubbard Street
Suite 201
Chicago, Illinois 60654
United States
+1-312-477-7300

London

167-169 Great Portland Street
Floor 5
London
W1W 5PF
United Kingdom
+44 (0)20 7834 5000

INQUIRIES

[email protected]

Ready to learn more?

schedule a demo