Text Analytics Newsletter #2

Exclusive E-roundtable discussion

Hi

Please find below your free copy of the brand new Text Analytics News newsletter. Providing you with the most timely news about the market. If you would like to be a guest editor, simply contact josh@textanalyticsnews.com

Like what you read and would like to receive more from Text Analytics News, simply click on the tab on the top right hand side of the page and sign up today - for FREE! If you do, you will receive:

• Regular updates on the online video search market
• Expert advice on a wide range of issues that impact your business
• Interviews with industry experts

Text Analytics News is also pleased to announce the recent release of our exclusive whitepaper, authored by Seth Grimes President of Alta Plana corp entitled "Text Technologies in the mainstream". Simply click on the tab on the right hand of the page to download it for FREE!

If you know anybody else who might be interested, then forward this newsletter to them!

Don't forget to subscribe today!

Many Thanks

Joshua Bull
Director
Text Analytics News
www.textanayticsnews.com
+44 (0)20 7375 7227

Hello and welcome to this exclusive Text Analytics News e-roundtable discussion.

Taking part in today's discussion we're pleased to welcome

• Tom Anderson, Founder of Anderson Analytics
• Seth Grimes, President of Alta Plana Corp
• Olivier Jouve, Vice President, Corporate Development, SPSS
• Alessandro Zanasi, Security Advisor for the European Commission
• Justin Langseth, President and CTO, Clarabridge inc

Hosting the discussion was Joshua Bull of Text Analytics News

Tom: I think Text Analytics is the most exciting thing to happen to my field market research/analytics in quite some time. Text Analytics flips the 80/20 rule around for us analytics professionals. Prior to Text Analytics we had to focus on the 20% of data that was "structured" (quantitative/numeric), now we have the tools to work with the other 80% of the data which is far richer in many ways.

Seth: Text analytics is simply the use of software to automate document processing and to extract meaningful "features" - names, facts, attitudes - from text in order to turn documents into data sources. Documents as used in this definition range from e-mail messages and contact-center notes to on-line blog and forum postings to business reports and scientific papers with lots of other types in between. If it's text, regardless of language, and it contains business or other value, text analytics can help unlock that value.

Olivier: Text Analytics enables organizations to explore the "unstructured" information contained in text in much the same way that data mining explores tabular or "structured" data. Through text analytics, customers can uncover hidden patterns, relationships, sentiments and trends resulting in greater insight from articles, surveys, call center notes, e-mails, blogs, and other types of text documents. Organizations using text mining easily extract the predictive power from these channels to reduce customer churn, improve productivity, detect fraud and increase marketing campaign results.

Alessandro: Text Analytics is a branch of Artificial Intelligence and regards all the activities which simulate the human capabilities in managing and analyzing the text. So NLP, Information Extraction, Text Mining and several other text disciplines are included in TA. Thanks to TA we can retrieve, analyze and synthesize 1 million of Multilanguage web pages, transcripts of phone calls, emails, blogs...extracting the "important" facts for the reader, giving him a synthesis in some lines.

Justin: Text analytics gathers textual information from a company's internal and external sources and structures the data to create useful customer experience intelligence. Sentences and paragraphs from unstructured content are automatically transformed into reports, calculations, and presentation views that can be easily analyzed. Virtually unlimited volumes of data can be quickly processed to discover relevant trends and potential issues and provide a holistic view of how a company is operating at every level.

With text analytics, all of the previously unusable customer feedback is available alongside existing structured, transactional customer information, such as purchase history, demographic information, product information, geographic information, etc. Text analytics is the key enabling technology for CEM. Text analytics enables companies to glean a true 360-degree view of customer desires, emotions and future behaviors, and ultimately, quantify the voice of the customer.

Tom: The first companies to benefit are likely to be the fortune 1000 that can afford the time to look into the new technology or hire consultants to do it for them. But text analytics applications will soon be added to more web applications and many people will benefit from the insights. This has already started happening.

Olivier: Around 80 percent of data stored within a company is in the form of text documents, which is the key factor in enabling a company to gain a complete understanding of their customers' behavior and attitudes. So, any company, across any industry including the public sector, who wants to derive better knowledge of their employees, customers or constituents can reap great rewards from using Text Analytics to improve business processes and customer satisfaction. For instance, Nucleus Research, a leading IT research analyst firm, found that text mining technology can reduce customer churn by more than 50 percent when properly implemented as part of an overall customer satisfaction strategy.

Seth: Certain enterprises generate and/or consume vast amounts of textual material. These include long-standing text-analytics users such as the life sciences and intelligence as well as more recent adopters including the legal and financial sectors. Text analytics is huge in media and publishing. But beyond verticals, it can benefit just about any company that makes, markets, or sells products and services to consumer markets. These organizations can benefit from Voice of the Customer initiatives that tap sources ranging from surveys to contact-center transcripts to on-line blogs and forums. These initiatives provide extremely valuable insights for customer care, sales, marketing, and product design.

Alessandro: All the companies which, for their objectives, need manage large quantities of text. If, in the first years of Text Analytics, the principal companies were in Pharmaceutical (for Competitive Intelligence) and Media (for Knowledge Management), in the last years they are in Services (e.g. Banking, for Customer Services applications). Now there is a strong interest, expected to grow in the next years, in Security (either in government/law enforcement organizations and in private companies offering technologies and services to government).

Justin: Customer-centric organizations, those with large numbers of customers and large numbers of products or services, find text analytics the key enabling technology for their customer experience initiatives. A sampling of these industries include: Consumer Packaged Goods; Financial/Insurance Services; Pharmaceutical/Life Sciences; Hospitality; Entertainment; Marketing Services; Retail; Healthcare; Investigation; Government.

Tom: It's quite amazing. Last year no one knew what Text Analytics was, now companies are calling me and say they need to know what text analytics can do for them. I wouldn't say the technology has changed all that much over the past year, rather the way we use it has changed. We're now much more adept at leveraging the technology to solve various types of business problems.

Seth: Rapid expansion of the scope of applications and of market acceptance have driven technology innovations in the last year. I'll single out one direction: very significant attention to extraction and analysis of attitudinal information, of opinions and sentiments expressed in social and traditional media

Olivier: Text Analytics has seen improvement in four main areas:

1. Broader understanding of customers' perspectives, taking into account qualitative feedback by extracting and classifying unstructured data in multiple languages and formats (written and audio).
2. Better clarifies customer feedback through sentiment extraction to automatically analyze and score patterns of positive and negative opinions.
3. More deeply integrated with Predictive Analytics to ensure one analysis methodology, with seamless deployment and a proven ROI.
4. Easier and more available for business users and technical analysts.

Alessandro: Technology has especially changed in developing advanced solutions for largely diffused exotic languages: e.g. Arabic and Chinese. Showing interest for limited diffused languages: e.g. Farsi, Pashto, and Dari.

Justin: The past year found Text Analytics maturing to be useful for general users instead of linguistic or analytical experts because:

1. BI Tools have all reached high version numbers and achieved maturity in terms of both end-user functionality and enterprise deployment.
2. Data Warehousing. Enterprise data is now centrally accessible for analysis and action. Data warehouses have dramatically increased to multi-terabyte sizes, and are now ready to accommodate a voluminous influx of unstructured data.
3. CRM: Enterprises have invested heavily in CRM automation to collect large amounts of data on customers.
4. NLP or "natural-language processing", converts human language into representations that computer programs can manipulate. NLP has been integrated as a capability within text-mining platforms, relieving enterprises of the complexities of linguistics training and other implementation barriers that plagued early technology solutions.

Tom: The extraction and coding is becoming a commodity to some degree. Naturally there are differences in products. But I believe the true benefit in terms of insights will come from early adapting consultancies such as Anderson Analytics which have analytics knowledge beyond just software programming. Engineers and IT professionals have been using the technology up until now. They are smart, but they are not used to thinking as analytics consultants/market researchers, so they have a hard time turning the data into actionable insights. This is specifically our strength!

Seth: What' most exciting is seeing text analytics come into its own: the recognition both from BI and predictive analytics practitioners and from business executives and analysts that they can and should be applying text solutions to expand the scope and power of their analyses.

Olivier: The increasing importance of analyzing social media - consumer blogging, social networks, community boards, and wikis - is changing how firms measure marketing effectiveness. Catching the Voice of the Customer through Surveys and Web 2.0 data through sentiment analysis, and coupling that with existing structured data, such as demographics and transactional history, permits more accurate results, better predictive modeling and customer understanding.

Alessandro: I really find an exciting thing the fact that governments all over the world decided of starting to fund the research and development in this new field: the results of this research will allow the prices for strongly advanced solutions to drop down drastically, allowing a real enlargement of the market.

Tom: The extraction and coding is becoming a commodity to some degree. Naturally there are differences in products. But I believe the true benefit in terms of insights will come from early adapting consultancies such as Anderson Analytics which have analytics knowledge beyond just software programming. Engineers and IT professionals have been using the technology up until now. They are smart, but they are not used to thinking as analytics consultants/market researchers, so they have a hard time turning the data into actionable insights. This is specifically our strength!

Seth: Text analytics will track IT developments with wider adoption of "as a service" hosted/on-demand models that make it easier for companies to get into the technology. In addition, we'll see the technology increasingly embedded in business intelligence and predictive analytics applications and also in media and publishing applications where the users won't even know they're doing text analytics. And text analytics will continue to make search smarter through more capable and wide-spread natural-language interfaces, deeper semantic capabilities, and through "question answering" systems.

Olivier: Anything is possible. Blog mining is a bit old news. We've been especially interested in combining text analytics with link analysis to understand social network services (like Facebook and LinkedIn), but text analytics will have far reaching implications and benefits.

Alessandro: I see a movement towards integration and consolidation: different technologies are integrating to deliver more complete solutions; big companies are looking for SMEs with some "unique" or start-ups with particular capabilities to be bought and merged together.

Justin: Fortune 1000 companies have embraced text analytics as the means to listen to, analyze and act on and measure their customer feedback. As most of those companies have constituencies in other countries or require support with foreign language for their North American operations, it seems extremely likely that multiple-language support will be needed.

I'd like to thank our e-roundtable industry experts for contributing their insight and expertise to the discussion. If you'd like further information please sign up to our weekly newsletter, and visit our website for the latest news and developments in the Text Analytics space at www.textanalyticsnews.com

We'd also like to invite you to download our exclusive whitepaper authored by Seth Grimes President of Alta Plana and summit chair of the 4th annual Text Analytics Summit, entitled "Text technologies in the IT mainstream" please click on the link to download it now for free! www.textanalyticsnews.com/4thannual08/whitepapers.shtml

Best

Joshua Bull
Director
Text Analytics News
josh@textanalyticsnews.com