New search engine launch is a great deal, especially when it is based on ideas, that were never implemented in search before. Wikia Search, that was launched recently is the first opened search engine, based on principles of transparency, collaboration, quality and privacy.
Wikia Search is developed under Wikia.com, that was founded by Jimmy Wales, who is also known as Wikipedia founder. Due to the great interest of our readers to all brand-new search engines, we decided to take an interview with Angela Beesley, co-founder of Wikia.com and Jeremie Miller, Technical Leader of Wikia Search. Fortunately, they agreed to answer our questions :)
- Wikipedia is one of the most popular sites, why the idea of new searchengine
has appeared? Isn’t it too risky to come to brand-new market?
Angela: Although both Wikia and Wikipedia were started by Jimmy Wales, they are totally separate companies. The organization that manages Wikipedia is not involved with Wikia Search.
Wikia is a commercial wiki company. As well as developing Wikia Search, Wikia also hosts 4000 wiki sites on any topic a group of people are passionate about. These range from gaming to pets to politics.
Here is a selection of the Russian wikis we host:
More than 1000 articles
100 to 500 articles
50 to 100 articles
All of these wikis are open to editing, and anyone can request a new wiki if their favorite topic isn’t covered yet.
- What technology is the basis of Wikia Search? Was it your own development, or you have taken some open-sourced products? What software and hardware do you use?
Angela: All of the software is Open Source. We are building on the work of many open source projects, including Nutch, as well as developing our own software which will be freely licensed and available for anyone to use.
Jeremie: It’s based on the open source projects Nutch (for search) and Foowi (for social). There’s over 2400 CPUs hosted at the Internet Software Consortium (ISC.org) but only a couple hundred or so in use at any one time right now.
- How much storage do you have? How many documents are already indexed
Jeremie: We have over half a billion URLs in our database, but only between 10 and 20% cached at this time.
- How many people are involved in this project? Are they all volunteers, or
you have some hired personnel? Was it hard to attract volunteers for this
Angela: This is primarily a volunteer-driven project.
Jeremie: There are a couple hired people, but mostly volunteers. Over 10 thousand have registered on the social side, and over a thousand active across the mailing lists, wikis, and related open source projects.
-What are the principles of ranging? Is it entirely human-powered, or
recombinant with machine algorithms?
Angela: We intend to combine the two. We will provide human-created results, and also each user’s social graph will be able to influence their results. The results you see will eventually be affected by the actions of the friends you have in the social networking part of Wikia Search.
Jeremie: Mostly plain Nutch with OPIC right now, but using the MiniArticles gives a human-powered tool with immediate impact.
- Will WikiSearch develop “natural language” searches?
Jeremie: Yes, and it will all be open source.
- Wikipedia is widely known as non-profit project without any advertisement,
will WikiSearch follow the same ides? How you will monetize you audience?
Are you planning to develop something like Google AdSense?
Angela: A couple of notes - this isn’t related to Wikipedia, and it is Wikia Search (not WikiSearch - that’s something else).
We’re focused on building Wikia Search first. After that, we will think about monetization! We will probably use advertising, though Google AdSense might not be entirely appropriate for this considering Google Search is a competitor.
Jeremie: There are many parts to the project that are free and open, but ultimately there will someday be some advertising similar to adsense on the search results from wikia.com.
- Human powered search is great idea, but aren’t you afraid of spamming
through it? What anti-spam mechanisms do you use?
Angela: As many large wikis have shown, if you have a good community behind the project who believe in your goal, then it is easy to defeat the spammers. We will use both automated methods and human effort to mitigate the problem of spam. With human-influenced results, I expect the spam problem to be much less than in would be with traditional search engines.
- How do you think, can Wikia Search gain same popularity in its market, as
Wikipedia? If yes, what time do you think it can take?
Angela: Wikipedia is a project to build an encyclopedia, not a search engine, so it’s hard to compare the two. I think Wikia, including Wikia Search, has the potential to be much larger than Wikipedia.
- Would you integrate Wikia Search to Wikipedia, for users who are searching in
Wikipedia, or something like that?
Angela: Wikia Search is open source, so Wikipedia are welcome to adopt our technology if they find that useful, as is any site which needs an internal search engine.
- When Russian, Ukrainian, Belorus interfaces will be launched? Are you
working on it now?
Angela: We hope to start launching international versions of Wikia Search later this year.
Editors of Searchengines.ru want to thank Angela and Jeremie for information and time and to wish their project Wikia Search to be a great success!
Written by Hanna Stus’
This article is also available in Russian