Megite is another news aggregator like Techmeme. Unlike Techmeme, however, it features different channels, and quite often it picks up on news items that I’ve missed on different feeds and aggregators. From time to time, I also notice it sends me a barrage of traffic when I get listed on the site as well. So, I thought it would be interesting to chat with the only name that is associated with the site, Matthew Chen, on what Megite is all about, how to pronounce the name, and where its going.

Tony Hung: I noticed your site some time ago as I started to get listed on Megite, but there’s not a whole lot to find on Megite about what it’s about and who the people behind the scenes are. Can you tell us a little bit about what Megite is, where you guys are located and a little bit about yourself?

Matthew Chen: We have been working on Megite related technology and algorithm for a long time for the other applications. But we’ve found that our algorithm is good at meme tracking, so we started to port our system and algorithm at the beginning of this year. We launched our site in February to test the waters. We’ve gotten lots of feedback, both in terms of suggestions and complaints. So we continue to improve it, and recently we’ve made significant improvements on the algorithm, software and infrastructure.

We want Megite to be an intelligent RSS reader. We create and develop the main verticals, such as technology, politics, business, entertainment, science, etc. And users can create their own Megite by feeding their reading list. The personalization service is costly and we are in the process of securing resources before we can launch it to the public.

We are located in Silicon Valley. Currently we have two people working on this project, both with computer science degree and many years of working experience.

TH: Does this mean you might be searching for external (i.e. VC) funding?
MC: Yes. We are going to approach VC’s for that purpose. If we don’t expand to personal version, we are ok to self funding.

TH: Where did you come up with the name “Megite” — and how do you pronounce it?
MC: This is a funny name originally for our other product.

Me - Mega(byte)
Gi - Giga(byte)
Te - Tera(byte)

me ‘gi te

me as in melissa
gi as in giga
te as t

TH: Can you talk a little bit about the most recent improvements? How will they make the experience better for your users?
MC: Sure.

1. New blog discovery: Introduce the blog auto discovery. So not only the A-list bloggers, B, C bloggers can be the sources to Megite. This improves the coverage and inclusiveness of Megite.

2. Ranking algorithm. Introduce the blog quality importance evaluation algorithm and improve the ranking algorithm by combining the blog importance, content relevance, linkage information, time, user feedback, and more.

3. Clustering algorithm. Better clustering algorithm to put the posts about the same topic under the same thread.

4. Better Thumbnail image extraction

5. Cache thumbnail images on our server.

6. Crawling infrastructure improvement that helps us to update more frequently and handle more verticals.

7. More verticals are covered: Entertainment, Science, Food, Fashion, etc.

8. Better title and description extraction. For example, the first release in Feb and the current version are very different. You can see better titles, better descriptions, better threads (cluster), better image thumbnails, and it’s more relevant. You can uncover the important news and at the same time balance the “new” news.

TH: Now by the blog discovery — is that something where bloggers can submit their own blogs? Or by that do you mean that Megite is simply looking for blog sources automatically?
MC: Mainly Megite is simply looking for blog sources automatically. Either they are mentioned by some good bloggers, or from with good and relevant content. Of course, a blogger can submit his/her blog to us and then they may be feed into our algorithm if it is not a spam blog. A blogger can link to us or news items, and if the algorithm sees lots of references to that blog, it may pick up the blog and feed into the algorithm.

TH: So, what you’re saying is that while most of the time Megite picks up blogs on their own because of link patterns and relevant content, bloggers *can* submit their sites to be included?
MC: Right. For a quality blog, Megite can pick up automatically. For the newly created blogs, a user can submit to us. We will evaluate the quality. Otherwise it has to wait for our algorithm to evaluate in the future.

TH: How do you see Megite different than other news aggregators such as Techmeme or TailRank? (Besides that it doesn’t start with a “t”)
MC: We believe the differences to be in terms of Quality, Coverage, Comprehensiveness, Relevance, and the fact that it’s up to date. We have been working hard to put ourselves ahead.

Techmeme and TailRank came to market before us. But we are in good shape now. We still have lots of room to improve. The first difference is that our powerful algorithm makes it very easy to create new verticals.

Because of the auto discovery, Megite can keep the pace with the blogspace. Normally good high quality blogs can be auto discovered.

Also, Megite is able to deliver the personalized version with the user’s reading list.

TH: Anything we can look forward to in the future?
MC: First a comprehensive news and blog aggregation portal and platform. Megite will cover news and blogs on every aspect of life, tech, politics, science, life, art, music, etc. Second a personalized, customizable news reader that works on user’s reading list. Will be able to recommend relevant blogs in the future. Third, search on the archives