Bitext Delivers a Breakthrough in Localized Sentiment Analysis
May 29, 2013
Identifying user sentiment has become one of the most powerful analytic tools provided by text processing companies, and Bitext’s integrative software approach is making sentiment analysis available to companies seeking to capitalize on its benefits while avoiding burdensome implementation costs. A few years ago, Lexalytics merged with Infonics. Since that time, Lexalytics has been marketing aggressively to position the company as one of the leaders in sentiment analysis. Exalead also offered sentiment analysis functionality several years ago. I recall a demonstration which generated a report about a restaurant which provided information about how those writing reviews of a restaurant expressed their satisfaction.
Today vendors of enterprise search systems have added “sentiment analysis” as one of the features of their systems. The phrase “sentiment analysis” usually appears cheek-by-jowl with “customer relationship management,” “predictive analytics,” and “business intelligence.” My view is that the early text analysis vendors such as Trec participants in the early 2000’s recognized that key word indexing was not useful for certain types of information retrieval tasks. Go back and look at the suggestions for the benefit of sentiment functions within natural language processing, and you will see that the idea is a good one but it has taken a decade or more to become a buzzword. (See for example, Y. Wilks and M. Stevenson, “The Grammar of Sense: Using Part-of-Speech Tags as a First Step in Semantic Disambiguation, Journal of Natural Language Engineering,1998, Number 4, pages 135–144.)
One of the hurdles to sentiment analysis has been the need to add yet another complex function which has a significant computational cost to existing systems. In an uncertain economic environment, additional expenses are looked at with scrutiny. Not surprisingly, organizations which understand the value of sentiment analysis and want to be in step with the data implications of the shift to mobile devices want a solution which works well and is affordable.
Fortunately Bitext has stepped forward with a semantic analysis program that focuses on complementing and enriching systems, rather than replacing them. This is bad news for some of the traditional text analysis vendors and for enterprise search vendors whose programs often require a complete overhaul or replacement of existing enterprise applications.
I recently saw a demonstration of Bitext’s local sentiment system that highlights some of the integrative features of the application. The demonstration walked me through an online service which delivered an opinion and sentiment snap in, together with topic categorization. The “snap in” or cloud based approach eliminates much of the resource burden imposed by other companies’ approaches, and this information can be easily integrated with any local app or review site.
The Bitext system, however, goes beyond what I call basic sentiment. The company’s approach processes contents from user generated reviews as well as more traditional data such as information in a CRM solution or a database of agent notes, as they do with the Salesforce marketing cloud. One important step forward for Bitext’s system is its inclusion of trends analysis. Another is its “local sentiment” function, coupled with categorization. Local sentiment means that when I am in a city looking for a restaurant, I can display the locations and consumers’ assessments of nearby dining establishments. While a standard review consists of 10 or 20 lines of texts and an overall star scoring, Bitext can add to that precisely which topics are touched in the review and with associated sentiments. For a simple review like, “the food was excellent but the service was not that good”, Bitext will return two topics and two valuations: food, positive +3; service, negative -1).
A tap displays a detailed list of opinions, positive and negative. This list is automatically generated on the fly. The Bitext addition includes a “local sentiment score” for each restaurant identified on the map. The screenshot below shows how location-based data and publicly accessible reviews are presented.
Bitext’s system can be used to provide deep insight into consumer opinions and developing trends over a range of consumer activities. The system can aggregate ratings and complex opinions on shopping experiences, events, restaurants, or any other local issue. Bitext’s system can enrich reviews from such sources as Yelp, TripAdvisor, Epinions, and others in a multilingual environment
Bitext boasts social media savvy. The system can process content from Twitter, Google+ Local, FourSquare, Bing Maps, and Yahoo! Local, among others, and easily integrates with any of these applications.
The system can also rate products, customer service representatives, and other organizational concerns. Data processed by the Bitext system includes enterprise data sources, such as contact center transcripts or customer surveys, as well as web content.
In my view, the Bitext approach goes well beyond the three stars or two dollar signs approach of some systems. Bitext can evaluate topics or “aspects”. The system can generate opinions for each topic or facet in the content stream. Furthermore, Bitext’s use of natural language provides qualitative information and insight about each topic revealing a more accurate understanding of specific consumer needs that purely quantitative rating systems lacks. Unlike other systems I have reviewed, Bitext presents an easy to understand and easy to use way to get a sense of what users really have to say, and in multiple languages, not just English!
For those interested in analytics, the Bitext system can identify trending “places” and topics with a click.
Stephen E Arnold, May 29, 2013
Sponsored by Augmentext
Analytics Company to Disrupt Digital and Mobile Metrics Emphasis
May 27, 2013
From Business Insider comes news of a potentially disruptive startup: “Mixpanel, A Startup That Wants To Kill Pageviews And Other ‘BS Metrics’ Now Measures 12 Billion Actions Per Month.” Mixpanel Co-founder Suhail Doshi pushes for digital and mobile companies to highlight monthly user engagement numbers instead of page views.
Mixpanel is an analytics company founded in 2009. It helps both paying and non-paying customers track engagement through actions on their sites. For example, “liking” content on Facebook is an action.
According to the article:
“Doshi admits it’s harder for content-producers to shift to his way of thinking. But changing an industry standard like pageview reporting is a slow process, and Doshi thinks his company is making good headway. ’We’re this living, breathing case that we do see pageviews are dying,’ says Doshi, who was inspired to track meaningful analytics by mentor and former colleague, Max Levchin. Pageviews are already dying on mobile devices, says Doshi, because users rarely click through to see more pages on tiny screens.”
Mixapanel’s growth implies they are doing something right. However, regarding Google Analytics, Mixpanel is making some bold assertions.
Megan Feil, May 27, 2013
Sponsored by ArnoldIT.com, developer of Beyond Search
Guide on Onboarding Funnel and Jargon
May 24, 2013
A recent article from Woopra caught our attention: “How To Build And Optimize An Onboarding Funnel.” This post explains what onboarding funnels are and how to utilize them to their fullest capabilities. The onboarding funnel is one of the key analytics reports for any SaaS company. According to this article are only 3 main steps to building and optimizing: tracking milestones, identifying major drop offs and optimizing problem areas.
Using Pinterest as an example, the article explains tracking onboarding milestones: this details the processing from signing up to pinning a first item. Pinterest would be able to see their major drop off at the step where users should follow 5 boards.
As far as fixing the problem areas, the article suggests looking more granularly at the problem and identifying the cause:
“Sometimes even going so far as to split up one step into two or more can help you diagnose the cause of a problem. For example, if you notice many users begin filling out your signup form, but then abandon it, you may want to separate the different sections of the form into several pages in order to see which section is causing users trouble. You may very well find that it is the requirement to add credit card information to start a free trial that is causing users to abandon.”
This article speaks to a common problem but instead of breaking things down to be more simple, the author of this post overlays a framework and appears to be the analytics jargon prize winner.
Megan Feil, May 24, 2013
Sponsored by ArnoldIT.com, developer of Beyond Search
Google as the Authority for Mobile Content
May 13, 2013
I my Google monographs, now out of print, I covered some of the early systems and methods Google developed to rank and identify “good” content. Now keep in mind that “good” is not the Manhattan Smith grad type of selectivity. Google’s “quality” processes involve mostly numerical recipes, data about who wants to advertise and for how much, and some “configuration” functions which give the algorithms some spunk.
I read “Google Launches Content Recommendation Engine for Mobile Sites, Powered by Google Plus”. The write up hooks the systems and methods to Google Plus, which is what makes sense. Google wants to make Google Plus a go-to social network, either crushing or buying such outfits as Facebook, LinkedIn or others. Google Plus is also working overtime to remain hip, timely, and relevant.
Information which is ready to heat and consume. No time consuming reading, analyzing, and evaluation. Image source: http://goo.gl/Tyy8a
What better way to achieve this that making Google Plus into the 21st century identify of what’s important and (more importantly) what’s not important. I think that those who think their content is important may have an opportunity to purchase some traffic, which is definitely supported by the Google infrastructure. Google is about revenue, not about objective search in my opinion. Your mileage may vary.
Here’s the passage in the write up I highlighted:
As Seth Sternberg, Google’s product manager for the Google+ platform told me last week, the team set out to create an “awesomely seamless experience to find more content” on the mobile web. On mobile sites, he argues, publishers often see high bounce rates because users have a hard time finding interesting additional content to read on a site once they have finished reading an article.
I like the use of “awesome” and I circled the “high bounce rates”. Yes, definitely an opportunity to deliver a better solution to mobile users. Those tiny devices just don’t delivery content and user access the way my three wide screen monitors and old-school, clicky Rosewill keyboard deliver the content bacon.
Several observations:
First, the application of Google recommendation technology to mobile is indeed a very significant step for the Google. No one expects humans to keep pace with the new content flooding the tubes of the Internet. The simplicity and appeal of “let Google do it” may make life tough for some folks.
Second, due to Google’s significant footprint in mobile, wherever Google goes has a significant impact. The mobile aspirations of outfits with fewer resources than Google are going to have to work overtime to make their business models hum. Online has a charming quick. Online services tend to form monopolies, squeezing out secondary and tertiary services the way weeds choke the trees next to the Harrod’s Creek post office, which still is open on Saturday.
Third, developers may just find it easier to embrace the Google. Yahoo is allegedly suffering a mild form of eczema from the Microsoft search deal. As Yahoo valiantly tries to deliver on the former Googler’s business strategy, Yahoo may just find it better, faster, and cheaper to open the door to Google’s walled garden.
Fourth, users who are now struggling to read above the 8th grade level may just take what the giant services deliver. Who wants to think about a query and then read a list of results. Once that’s done, the user bristles at the thought of opening documents, ingesting them, and then analyzing the content for the needed information.
Nope, just nuke that information burrito in the Google microwave. Recommended content is ready to consume. Very modern and very, very appealing to advertisers and those who will pay to be included.
Stephen E Arnold, May 13, 2013
Sponsored by Augmentext, the original flash frozen information burrito.
The Boston Marathon Bombing And Video Analytics
May 4, 2013
While the aftermath of the Boston Marathon Bombing continues on with the perpetrator’s trial and discovering how far his plans extended, the “Video Images Yield Two Possible Boston Bombing Suspects.” A department store camera caught one of the bombers leaving a backpack and then another camera caught the pair acting undisturbed by when the bombs went off, thus singling them out. The authorities relied on photographs and video evidence to piece together the events, which augmented the collected physical evidence. Already the disaster is being deemed the most photographed bombing in history. Authorities used sophisticated software with algorithms that can track patterns, i.e. clothing color and object movement.
“’The question that is most often asked is, is there a button we can push to make this happen as quickly as the general public thinks we can, from watching television and movies,’ said Larry Compton, operations manager at Forensic Video Solutions Inc., a firm that serves as a consultant to law enforcement. ‘The answer is no. These tools and techniques are really designed to focus the analysts,’ he said.”
As seen the technology does have its weaknesses, mostly due to low-resolution and lack of visual detail, which is why the authorities turned to the public for help because cell phones have higher video resolution. Analytic software has its weaknesses, but the collaboration between the public and private law enforcement worked together to exact justice.
Whitney Grace, May 04, 2013
Sponsored by ArnoldIT.com, developer of Beyond Search
Google and Prediction: Not Thin Gruel, Please
April 28, 2013
The Google religion continues to diffuse in the UK. I read “Google Searches Predict Market Moves.” Predictive analytics can be useful. Years ago I was on the Board of Directors of NuTech Solutions, one of the then world leading firms in predictive analytics. The methods of what I call fancy math are extremely useful. Applications range from identifying markets in which certain type of automobiles are likely to sell to figuring out when to cut back on power plant operation to reduce pollution are of great value. However, fancy math based on mereology or more familiar techniques have limitations. What’s interesting is that the limitations of fancy math are of less interest than the science fiction types of applications which catch the attention of folks who write about math as opposed to do math.
Here’s a passage in the write up which I noticed:
Google’s own researchers found that searches can track the spread of influenza and more recently showed that they “predict the present”with regard to economic indicators. In 2011, the Bank of England determined that searches for relevant terms could even predict house prices.
Useful indeed. Now here’s the snippet I circled with my yellow highlighter:
The researchers have already been approached by executives within the financial industry to try to put their findings to use, and have recently received a grant from the Engineering and Physical Sciences Research Council to develop a “big data” software platform specifically aimed at the emerging business models that will depend on it.
The idea is that Google and those in the know can make money, save lives, and in general just do good stuff is compelling.
I would point out that Google has a number of systems and methods which make use of fancy math, Google generated metadata, and data provided by users and computer processes. These systems and methods are not new to Google. It would have been gratifying to some Googlers to have their work recognized. I think a reference to the work of Ramanathan Guha and Alon Halevy would have moved the write up from “gee, this is hot now” to a more foundational approach.
The write up would have gained impact in my opinion with a reference to Google’s support of the systems and methods of Recorded Future. Founded by the person who developed Spotfire, now owned by Tibco, is a cutting edge approach to generating actionable output from a range of data, including stock market related information. I can read the BBC story as if the information were based solely on Recorded Future’s capabilities. Note that Recorded Future has, in my opinion, pushed beyond some of its direct competitors like Digital Reasoning, Talend, and other firms in this market.
Yes, Google has formidable fancy math capabilities. I would be a happier old observer if those writing about fancy math included more substance in the write ups about Google. The thin gruel does not do Google justice nor does it present the scope of Google’s capabilities in an informed, rich context.
Stephen E Arnold, April 28, 2013
Sponsored by Augmentext, which also uses some fancy math
Predictive Application Tracks Global Events
April 27, 2013
Have you heard of the Global Data On Events, Location, and Tone project yet? Head on over to Foreign Policy and its article, “GDELT: What We Can Learn From The Last 200 Million Things That Happened In the World?” The article summarizes how the GDELT project tracks political events from all over the world. Similar databases exist for a particular region, but GDELT separates itself out by covering the expanse of the globe. It records events and categorize them by four different types: material conflict, material cooperation, verbal conflict, and verbal cooperation and within those categorizes the event is classified with CAMEO, a 300 category taxonomy system.
GDELT can be used to track political events and political rhetoric and from its data it can possibly predict the future and it might even be a tool for complexity theory mathematicians.
“Of course, for all the high-tech software behind its creation and its potentially far-out applications, GDELT is, at its core, a way of summarizing news coverage, and old fashioned legacy-media news coverage at that. The sources used to identify events include world news coverage from Agence France Press, the AP, BBC,Christian Science Monitor, New York Times, UPI, and the Washington Post, as well as a few more specialized outlets and Google News. Leetaru notes in his recent paper introducing the project that the increasing availability of news on the web has led to a ‘dramatic increase [of recorded events] since the beginning of the 21stcentury.’”
There are concerns for the project such as rural areas gaining as much frequency as developed areas and bringing in social media. Mainstream journalism has quality behind it, while social media is still relatively new and there is a lot of junk in it. The information needs to be gathered no matter where the source is from. The problem is sorting the wheat from the chaff.
Whitney Grace, April 27, 2013
Sponsored by ArnoldIT.com, developer of Beyond Search
Seek Inspiration from a Leading Web Content Expert
April 24, 2013
Web content and data analytics are surefire methods to improve the quality of a Web site as well as making more money for an organization. Connotate is a company that specializes in transforming Web content into powerful information assets. Connotate recently released “101 Reasons To Use Connotate,” highlighting the best the company has to offer. The list is broken down into “50 Ways Customers Use Connotate” and “51 Great Product Features.”
Perusing through the list it is obvious that Connotate can do many beneficial things, including saving time and money, streamlining processes, anticipate market trends, regulation compliance, web data monitoring, and creating new and innovative information products. Connotate fully endorses it products and services repeatedly:
“Connotate’s patented technology scales faster and easier than scripts or toolkits, saving you time and money. It is much more resilient to Web site changes, reducing maintenance costs. A built-in management layer provides workflow support to boost operational productivity.”
Connotate’s list is quite remarkable in promoting it services. Data monitoring is a growing industry and plays into the big data trend, but Connotate actually tells to you what can be done with data. Other companies are going to turn to this list to improve their own offerings and services.
Whitney Grace, April 24, 2013
Sponsored by ArnoldIT.com, developer of Beyond Search
Hadoop in Demand Yet Lacks Trained Professionals
April 24, 2013
Hadoop has been in the headlines lately for its major changes and how it is being integrated into more organizations. PR NewsWire takes a look at the open source database platform and what it predicts will happen for the company in the future in, “Global Hadoop Market 2012-2016-Lack Of Trained Professionals To Be A Major Challenge.” The article examines a recent TechNavio report that analyzes the Global Hadoop Market 2012-2016. TechNavio predicts Hadoop will grow at a CAGR 55.63%, mainly due to rise in big data analytics and the company offering Hadoop-as-a-service. While technology and service wise Hadoop is doing well, it faces a deficit in trained professionals who can do the work.
TechNavio said:
“’The demand for cost-effective Hadoop-based big data solutions is driving this market. Organizations understand the importance of big data solutions, but installing and hiring new professionals to deploy them is a costly affair. As a result, organizations and decision makers are adopting Hadoop-as-a-service (HDaaS) solutions that provide cost-effective big data management and analytics. HDaaS solutions offer the necessary hardware, software, and services required to support big data management at low subscription fees.’”
What seemed to be a straight shoot, Hadoop is facing a problem that might limit its growth and development. HaaS does take care of part of the problem, but someone still has to work with the software. Will Hadoop innovation shift to where the proficient professionals are? We think it is a strong possibility.
Whitney Grace, April 24, 2013
Sponsored by ArnoldIT.com, developer of Beyond Search
The Evolution of Analytics
April 23, 2013
In the Kontagent Kaleidoscope article “Analytics: 2012 Is For Mobile What 1997 Was For The Web” marketing executive provides a video overview of the evolution of analytics. Known as an expert in his field, Schulman makes some bold statements such as the Web is dead. Schulman believes that in its current state mobile marketing is the same as simply measuring the number of hits on the Web. This does provide you with some information but it doesn’t give you everything you need. App developers try to get their apps at the top of user lists by testing names, categories and other things that might hopefully lead to the best rankings. He states that developers and marketers need to also be able to test the messaging, channels and segments to determine which of them would best attract potential users. In the beginning the Web only looked at visitors and not the unique users and there were only two kinds of marketing, direct mail and response. However, once Web analytics came along this also brought a whole new way to measure marketing and taught marketers how to not only track but also optimize campaigns.
“The evolution of marketing continues as we enter the multiscreen world. Think of it this way: 2012 is for mobile what 1997 was for the Web. Take it a step further; Kontagent is to mobile what Webtrends was to the Web. You can draw parallels around early adopters of Web analytics: these tools require some scientific methods of controls, A/B testing, cohort analysis, etc. Marketers need to take these proven methods and transfer them to measuring mobile users and in-app behaviors. In order to optimize conversions in mobile, imperative to help optimize conversions in mobile.”
If mobile marketers dig a little deeper they can use analytics to figure out where they should be focusing their mobile marketing for optimum results. Remember some things are more than just skin deep.
April Holmes, April 23, 2013
Sponsored by ArnoldIT.com, developer of Augmentext
-
- Subscribe to Beyond Search
Feature archive
News archive
-
