|Have you ever wondering why modern search engines are relying only on a keyword search? Why there is no neat logical structure to search the web? Why there are tons of irrelevant results, just because occasional occurence of search phrase in some text, which logically is absolutely irrelevant to your request? Why useful information from regular users quickly get lost on the web just because the user have not much followers?
I was wondering about the same concerns for a long time. It was so long, so I finaly decided to deal the problem by myself. And now introducing experimental concept of knowledge base engine, Project Identity ( http://project-identity.com/ ) , which have chance to cope with above problems.
Its key features:
- each piece of information is organized into properties and relations, called knowledge
- search is performed with respect to such a structure. it is much more powerful than regular keyword search, avoiding bloating of results with occasional occurence of target keywords.
- any user have full control of which people to trust during search. no spam or other unwanted information. freedom of choice for any alternative opinions, disregarding their popularity.
- usefulness and precise logical structure are primary criteria making knowledge easy findable by other users
- each user have full control of his knowledge, no censure or moderation. knowledge will not disappear due to strict moderation or occasional downvoting.
- knowledge can be arbitrary small, describing narrow aspect of any phenomenon. complex picture of the phenomenon can be gathered from such small knowledges via proper search request.
The idea of the system, people are entering knowledge by themselves. They are entering anything they believe useful, and stating the properties, what they are believe crucial. This information may include any object properties, including its crucial features, possible search phrases to make it easy discoverable, some extended description, link to detail web page. Then these knowledge are shared between all the people, even not logged into the system. So, people are helping each other to find something, they are believe crucial or hard to find via regular search engine.
It is even possible to enter knowledge about some object, which is not presented on the web. For example contacts of your local hairdresser, who is not familiar with web. You may enter his address to make it findable by address, when searching hairdresser nearby, and some his contacts: name, phone, working time.
Next example, you want to find some free software by its purpose. Trying to find it via regular search engine may appear a real pain (especially it is less popular/advertised), because top results are commercial software and software catalogs. However having required keywords. With this system it is possible to specify software purpose, its license as a search criteria. Moreover, you can specify to find it with its developer site to have clear download link. The result will be not a mixture of keywords, but clear structured description corrsponding to the search request logic.
Next example. Imagine, there are audio-visual brain stimulation devices. You want to find some software for android that can produce both audio and visual brainwave pulsations. There are plenty of software with audio support, but rare with visual support (screen flickering as a replacement of audiostrobe glasses). Can you find it quickly via regular search engine? Tip: it is really existing.
The big problem of the web is to identify the information source person. All your know is the only profile, filled by person itself. It may lack info, be useless, or even wrong. If this is a conference, rating from conference owners and community may be biased.
Of course you may spend a lot of time tracking person activity: posts, discussions, multiple accounts on different sites or even discover the person in real life. But there is no efficient way to share this info to other people, or get such info from other people.
One of the main concepts of the system is the identity of the information source. All the information about person is collected together in its public rating and discussion. Anybody can add or see any information of the person. Even anonymous person have all its ratings collected together without revealing anonymity and separate from public ratings.
So for example it will be easy to discover who is speaking: well-skilled man or teenager trying to look tough. In regular web its often annoying to separate real man from soothers if you do not know them in real life.
Very close problem is absoluteness of different ratings and tops, which are formed by unfamiliar people and possible to cheat. Our system have a principal solution for the problem, each user is fully managing and switching any group of people, which will be trusted to rate search results.