With Sentiment Analysis Symposium coming up next month. I caught up with sentiment and text analytics expert Seth Grimes to talk about the state of the art, the upcoming conference, and what people can expect to get from it. I asked Seth four questions, here they are:
Seth, congrats on producing the 3rd Sentiment Analysis Symposium. Evidently sentiment is here to stay. Why is that?
Seth> Thanks Steve. We’re growing, reaching new audiences — the November 9 symposium is the first on the west coast — because folks from brand champions to policy-makers recognize business value in online and social sentiment.
There’s subjectivity in everything that happens online, 100%. Even a simple Foursquare tweet — “I’m at Philz Coffee w/12 others” — is more than a fact. It’s an implicit endorsement. It’s an invitation to promote. It’s an opportunity to forge or enhance a customer relationship. Maybe you can match the tweet to online opinions, posted to Facebook or Yelp. You aim to understand what customers — this particular customer -and the market as a whole really think about your products and services.
You aim to link those opinions to actual behaviors. Your goal is to measure, engage, and influence. Sentiment analysis is the key to doing the process right, to doing it quickly, accurately, and completely.
Sentiment analysis is evolving. What are the new directions and how will they be helpful to brands?
Seth> Sentiment analysis is going broad, it’s going deep, and it’s gaining options. It’s going broad in the sense of getting into new applications and into the hands of new users. We even have the Federal Reserve Bank of New York seeking to implement a Sentiment Analysis And Social Media Monitoring solution… as is just about any aware brand with a significant public face.
Sentiment Analysis is going deep in two senses. First, it’s jumping from a departmental solution, in the hands of just a few customer-support or market-research staff, into an enterprise solution with a key role in enterprise operations and decision making. Second, the technology is getting quite sophisticated. It’s getting at emotion in speech (importantfor contact centers), at “intent signals” that indicate opportunities and risks (key to online commerce and conversion), at support for precisely targeted semantic advertising. This stuff is far belong the positive/negative/neutral classifications that many people associate with sentiment analysis.
Brands have lots of options available to get at the information that’s of greatest value to them, to bring customer and influencer opinions and attitudes into play for just those business processes that are critical to their own success. The game hinges on targeting, competitiveness, effectiveness, profitability.
Even with progress, there are limitations to sentiment analysis. Where should brands be cautious in their use of sentiment analysis?
Seth> No question there are limitations, but I’m going to argue that the biggest ones relate to awareness. How do you choose the right technology? How do you use it effectively? What outcomes should you expect? How should you bake what you learn into everyday operations?
I’ll expand on one limitation I mentioned a moment ago: Low-grade tools that are capable only of rating tweets, status updates, and forum postings as positive, negative, or neutral, based on the presence of certain keywords. My goal professionally, and especially at the Sentiment Analysis Symposium, is to educate business users so they can make the bestchoices and make the most of what really is a very promising set of technologies.
What lessons and insights will attendees leave with?
Seth> Lesson number one involves understanding sentiment analysis as a part of larger business processes, aligned with business goals. You covered this in Listen First! — I’m thinking in particular of your Change the Research Paradigm chapter. Symposium attendees will hear directly from peers, analysts, and other experts how to go about it.
The variety of valid and valuable approaches will be a symposium revelation. Sure, we have talks about solutions that rely on automated natural-language processing (NLP), but we also have speakers who’ll cover crowd-sourcing and human sentiment analysis. The variety of applications – customer engagement, financial markets, retail, travel reviews, contact centers, real-estate markets! — will be another revelation.
I could say more, a lot more, but I’ll stop here and thank you for the opportunity to talk about sentiment analysis and the symposium, and I’ll also offer your readers a special registration discount, $100 off with the code ARF.
Related posts:
- 4 Questions for Seth Grimes, Sentiment Analysis Symposium Founder Seth Grimes, well-known authority on text analysis and sentiment, spoke with me about sentiment analysis...
- Learnings from Sentiment Analysis Symposium II, April 12 – UPDATE Seth Grimes with Listen First! The Sentiment Analysis Symposium and Listen First! both...
- Sentiment Analysis, Visionaries Panel, April 12 I’m thrilled to be moderating the “Visionaries Panel” at the Sentiment Analysis Symposium on April...
- 4 Questions about Sentiment and Text Analytics for Jeff Catlin, CEO, Lexalytics Jeff is appearing on the “Visionaries” panel at the Sentiment Analysis Symposium on April 12....

Pingback: Liens / Lectures 6 « TAL Etude de cas : Biocarburant