Search yesterday, today, tomorrow...
If I may start without the introduction and background.
Search engine today (including first and foremost, a web browser) is a software program which is based on the mathematical apparatus of statistical, probabilistic and other methods. In any case, he said. Considers links, considers the relevance of the statistics of transitions that takes into account many factors (location, age, etc., different situational information). This ultimately leads to a narrowing of results and the filtering issue. And that ultimately there is a huge, multi-level course and today is fundamentally quite a complex index to a database collected on the Internet information. In this case, the database information is also quite sophisticated, layered structure, which is understandable today, but does not change the fact. Here, of course, and caches, redundancy and parallelization, and the other, other, other, that provides each of us the opportunity to use, from my point of view, a very important resource. Just try to imagine today's Internet without searching. I'm even willing to argue that advances in information retrieval are the primary factor stimulating the growth of the Internet in principle.
However, what is search engine? The search engine is an intermediary between those who published and those who want to see published; between the thoughts of one man, converted into digital form some of the electronic document, and the thoughts of the other, presented in the form of a query. The search in this case, the communication channel with your communication Protocol, communication channel between in. This fact is extremely important: we're talking about the tool definitely enormous, but the tool of interaction of people in the vast majority of cases.
The other day I came across an article four years ago habrahabr.ru/post/31600 in which the problem, or rather the idea of semantic search, in this connection has any objections, questions, and answers.
1. The quality of search today. What is the level? What are the prospects?
Theoretically, the maximum achievable quality, based on current technologies, is when I request I get one most relevant to the query article-the answer! That is, given the maximum possible number of factors of mathematics machine search computes, such compliance. At the same time we must understand that the search engine will display the fact that someone left. Reaching this theoretical level of our communication channel (search engine), we ask the second question: how is that mathematically the answer - the answer from the point of view of the mind? Because the ideal response we can get, if the returned result was actually a response to someone specified exactly our issue. For my purposes the level of today's search is sufficient. So, I was quite comfortable and quick to find information I. Improving relevance in currently used architecture in General, as far as I know, is achieved by increasing the participating in the process parameters include in the request the maximum number of data available for greater differentiation of issuance.
2. Semantic search - what is it?
The content search or search with a sense? I will not argue about definitions, but searching for content with the understanding of the meaning is a completely different technology platform. It is a completely different architecture. Where the system acts as: "learn, understand, ask a question, form an answer, answer." All I see now is finding information format of question-answer, again, is communication of people. That brings the functions of a search engine is still the same math.
The problem lies in the field of my interest, research in this and related fields and results. We Bibikom, are the answer project, which tested the results obtained in the field search. However, the search is not the only that require different approaches sphere.
In doing so, I rethink a lot of things, until the very concept of information, principles of its organization, and processing. I don't like the idea of presenting information in a special machine-oriented form. It will not lead us to "smart" computer, but rather will require a lot of specialists-programmers, as happened with today's programming ( what I want to say separately ).
I am sure that the search tomorrow is already the person<->machine. Where the car is a completely different technology platform, for which the information will cease to be a senseless array of bytes. I would like not only to live up to these times, but to make the best of their efforts!
Article based on information from habrahabr.ru
Search engine today (including first and foremost, a web browser) is a software program which is based on the mathematical apparatus of statistical, probabilistic and other methods. In any case, he said. Considers links, considers the relevance of the statistics of transitions that takes into account many factors (location, age, etc., different situational information). This ultimately leads to a narrowing of results and the filtering issue. And that ultimately there is a huge, multi-level course and today is fundamentally quite a complex index to a database collected on the Internet information. In this case, the database information is also quite sophisticated, layered structure, which is understandable today, but does not change the fact. Here, of course, and caches, redundancy and parallelization, and the other, other, other, that provides each of us the opportunity to use, from my point of view, a very important resource. Just try to imagine today's Internet without searching. I'm even willing to argue that advances in information retrieval are the primary factor stimulating the growth of the Internet in principle.
However, what is search engine? The search engine is an intermediary between those who published and those who want to see published; between the thoughts of one man, converted into digital form some of the electronic document, and the thoughts of the other, presented in the form of a query. The search in this case, the communication channel with your communication Protocol, communication channel between in. This fact is extremely important: we're talking about the tool definitely enormous, but the tool of interaction of people in the vast majority of cases.
The other day I came across an article four years ago habrahabr.ru/post/31600 in which the problem, or rather the idea of semantic search, in this connection has any objections, questions, and answers.
1. The quality of search today. What is the level? What are the prospects?
Theoretically, the maximum achievable quality, based on current technologies, is when I request I get one most relevant to the query article-the answer! That is, given the maximum possible number of factors of mathematics machine search computes, such compliance. At the same time we must understand that the search engine will display the fact that someone left. Reaching this theoretical level of our communication channel (search engine), we ask the second question: how is that mathematically the answer - the answer from the point of view of the mind? Because the ideal response we can get, if the returned result was actually a response to someone specified exactly our issue. For my purposes the level of today's search is sufficient. So, I was quite comfortable and quick to find information I. Improving relevance in currently used architecture in General, as far as I know, is achieved by increasing the participating in the process parameters include in the request the maximum number of data available for greater differentiation of issuance.
2. Semantic search - what is it?
The content search or search with a sense? I will not argue about definitions, but searching for content with the understanding of the meaning is a completely different technology platform. It is a completely different architecture. Where the system acts as: "learn, understand, ask a question, form an answer, answer." All I see now is finding information format of question-answer, again, is communication of people. That brings the functions of a search engine is still the same math.
The problem lies in the field of my interest, research in this and related fields and results. We Bibikom, are the answer project, which tested the results obtained in the field search. However, the search is not the only that require different approaches sphere.
In doing so, I rethink a lot of things, until the very concept of information, principles of its organization, and processing. I don't like the idea of presenting information in a special machine-oriented form. It will not lead us to "smart" computer, but rather will require a lot of specialists-programmers, as happened with today's programming ( what I want to say separately ).
I am sure that the search tomorrow is already the person<->machine. Where the car is a completely different technology platform, for which the information will cease to be a senseless array of bytes. I would like not only to live up to these times, but to make the best of their efforts!
Комментарии
Отправить комментарий