Seth Grimes, president and principal consultant of Alta Plana Corp. and founding chair of the Sentiment Analysis Symposium, has put together a thorough new report, Text Analytics 2014: User Perspectives on Solutions and Providers. Among the interesting findings of the report is that “growth in text analytics, as a vendor market category, has slackened, even while adoption of text analytics, as a technique, has continued to expand rapidly.”
Grimes explains that in a fragmented market, consisting of everything from text analytics services to solution-embedded technologies, the opportunities for users to practice text analytics is strong, but that increasingly text analytics is not the main focal point of the solutions being leveraged.
Reflecting the diversity of options, respondents listed among their providers a number of open-source offerings such as Apache OpenNLP and GATE, API services such as AlchemyAPI and Semantria, and enterprise software solution and business suite providers like SAP. The word cloud above was generated by Alta Plana at Wordle.net to show how users responded to the question of companies they know provide text/content analytics functionality. Nearly 50 percent of users are likely to recommend their most important provider.
Where’s the analytics action? Social remains the hot spot, with 61 percent of responders citing blogs, 42 percent noting news articles, and just under 40 percent in each case mentioning comments on blogs/articles customer/market surveys and online forums.
It’s interesting to note, though, that not quite one-quarter are leveraging the technology for contact-center notes or transcripts – especially given the finding that top business applications of text/content analysis include brand/product/reputation management (38 percent), and voice of the customer/customer experience management (39 percent), and also considering some of the return on investment measurements customers have in place.
The survey reports that 19 percent of users are measuring ROI as it relates to higher satisfaction ratings; 21 percent as it relates to increased sales to existing customers; and 17 percent to higher customer retention and loyalty. Out of 170 respondents, 42 percent have achieved positive ROI according to some categorized measure.
According to the survey, the main capabilities users are looking for in a solution, each with over 50 percent selection, include the ability to generate categories or taxonomies; customizability in the form of the ability to use specialized dictionaries, taxonomies, ontologies, or extraction rules; broad information extraction capabilities; and document classification. Forty-five percent of respondents claimed a need for more specific information extraction – i.e., deep sentiment/emotion/opinion, compared to 57 percent in the 2011 survey. Sentiment scoring was noted by 41 percent of respondents as being an important solution feature.
That said, there’s room for improvement in a couple of important areas, as well Twenty-nine percent of users are disappointed or very disappointed in the technology’s ease of use, and 25 percent in results accuracy.