📜 ⬆️ ⬇️

Blippex - Search Wikipedia

Today, the search engine market is represented by quite a number of services, the leadership among which, of course, is what Google takes. Competitors, among which there are serious players such as Microsoft Bing, Yahoo, Yandex, DuckDuckGo, try to keep up with the race. But almost all of these systems have one thing in common - they are trying to imitate and duplicate Google algorithms. Someone pays attention to search personalization, someone cares about the anonymity and privacy of users, but everyone tries to provide almost identical search results.

Blippex - Wikipedia search, this is a new search engine, one of the few that are not trying to clone Google. We believe that in order to necessarily compete with search leaders, it is not necessary or even necessary to provide the same search results as other systems. We believe that the desire to show almost identical results has been a deterrent to the development of search engines in the last decade.



The Blippex search engine uses a different approach to creating a search index and, as a result, a different search and ranking algorithm. Our idea is that users themselves should provide search information. This approach is somewhat similar to Wikipedia, but unlike Wikipedia, our users should not do something manually, but can only install a special browser extension that will anonymously send information about the pages visited and help us build a search index and influence the importance of information.
')


We believe that the amount of time spent on viewing a particular page or site is a good indicator of the importance and relevance of information. We call this algorithm DwellRank. As you know, the Google search engine uses the PageRank algorithm, which is based on the analysis of the relationship of pages and sites to each other. Of course, that DwellRank takes into account not only the time spent by the user on the page, the indicators of the number of visitors, the length of the content, the popularity of the site as a whole are taken into account.

Blippex pays great attention to the anonymity and privacy of users. We do not collect or store any personal information about our users. All we need to build a search engine is only the address of the page and the time spent on it. By the way, the source code of the plugins is open, so that anyone can verify the data that the plugin collects. For example, Google search engine can not boast of anonymity, because the personalization of search results and advertising are the main business and income of the corporation. Moreover, every month we upload a dump of our database to free access.



More recently, we have taken another step towards greater anonymity - we have provided the peer-to-peer (peer-2-peer, WebRTC) network option for our extension under the Chrome browser. The main goal of the innovation is to ensure greater anonymity of the data on visiting any page. In our case, instead of direct sending to the server, the information will first pass through a random number of other clients, which will make it impossible to determine the original sender.

At the moment the service is at the beta stage, we have about 50 thousand unique visits every day and about 400-500 thousand search queries. Search index is 10 million pages. Some search queries all may still produce unexpected results, but we are working hard to improve our algorithms. Anyone can help us by installing our extension. The variety of data helps us to carry out better tuning of the search itself and, of course, to provide better results. In any case, the launch of the Blippex search engine spawned many lively discussions and discussions and made us clearly understand that such a product has a future, because people tend to trust what ordinary people create for ordinary people, while we are actively working to improve search algorithms and new functionality.

Source: https://habr.com/ru/post/199336/


All Articles