categories

HOT TOPICS

NEWSLETTER

If you are considering becoming a 1M/1M premium member and would like to join our mailing list to receive ongoing information, please sign up here.

Subscribe to our Feed

Thought Leaders in Big Data: Interview with Seth Redmore, VP of Product Management at Lexalytics (Part 2)

Posted on Sunday, Mar 17th 2013

Sramana Mitra: Let’s do a few use cases of your partners who are doing interesting applications using your base technology.

Seth Redmore: Are you familiar with a company called Bitly?

SM: Yes.

SR: So, what would a URL shortener want to do with text analysis? It turns out they have a very interesting view of social media, where they now own the content behind the links that people are shortening. That presents an interesting view to enterprises. We are part of their social media monitoring platform. The content that is behind that URL is passed through our system, and the “who,” “what” and “where” is being discussed inside of it.

SM: Can you talk a bit more about it? What we see as users of Bitly is that we stick in a URL and build a Shortnit to something – byte size, basically. What is the point of doing content analysis using your technology?

SR: The point is not for somebody like you and me. The point is for a company like Coca-Cola, so they can see in shortened URLs how often they are mentioned, if they are mentioned positively or negatively, what the context is in which they are mentioned, etc.

SM: I see. So that shortened URL being mentioned in social media is what you are helping with.

SR: Exactly. We are chasing what is behind that shortened URL. What is the content somebody actually wants to share? When you are out there acting on social media, Twitter has to make money, Bitly has to make money, etc. How are they going to do that? Both of them could take an advertising model, but the other way to do it is to take an analytics model. They can sell either sell content, like Twitter does, to various companies like Datasift – and I will talk about our relationship with Datasift in a second – and/or sell some kind of analytics platform. That is what Bitly has done. They said, “We are going to give you this unique view of social analytics of what pieces of content people thought were interesting enough to share and how these are reflecting on you Mr. Large Enterprise.”

SM: That is a good use case. Let’s do another one.

SR: At the other end of that pipe we are partners with Datasift. Twitter has only a certain number of companies they are going to give the fire hose to. They are basically Twitter’s license resellers. Datasift is one of them. They aggregate content from a lot of different places. They have Twitter, Reddit, Facebook, etc. They have a ton of things that are coming in. They also have a really nice API and language – this description language for describing “here is what I want.” For instance, “I want any content that mentions Toyota in the context of race tracks or McDonald’s in the context of food.” What we do is we help them provide a couple of things. Once of them is the sentiment for those things. If you see McDonald’s in the context of food and somebody says, “Great,” then you have positive sentiment. But we also help in determining whether it is in the context of food or not. That is as difficult as determining whether something is positive, negative or neutral. Somebody says, “I had a burger at McDonald’s.” That is food, because you know that a burger is food. But there are thousands of foods. So, we have technology that we developed, using a Wikipedia knowledge base, to have certain classes that they can break out like food, marriage and so on. They can classify content.

The other use is to determine what kinds of companies or people are being discussed. Maybe you are looking for something in a city. You know a city and you are getting all the content having to do with this city. As the city manager, you may be interested in what companies are being talked about with respect to a city, but you don’t have a list of the companies. Our engine will be able to determine, “Oh, this is a company.” It pulls it out and marks it as a company. Or, “This is a company, it is a company entity and it is positive for your particular city.”

This segment is part 2 in the series : Thought Leaders in Big Data: Interview with Seth Redmore, VP of Product Management at Lexalytics
1 2 3 4 5

Hacker News
() Comments

Featured Videos