KR20010108877A - Method For Evaluating A Web Site - Google Patents

Abstract

1. TECHNICAL FIELD OF THE INVENTION

The present invention relates to a method for evaluating a website and a computer readable recording medium having recorded thereon a program for realizing the method.

2. Technical problem to be solved by the invention

The present invention defines a property of a high quality site, and based on this definition, automatically evaluates each site on the Internet, thereby realizing a website evaluation method and method for providing users with high quality sites among numerous sites. Its purpose is to provide a computer readable recording medium having a program recorded thereon.

3. Summary of the Solution of the Invention

The present invention provides a method for evaluating a website applied to an information communication system, comprising: a first step of classifying a category to be evaluated; A second step of automatically extracting websites to be used for evaluation and classifying the categories; A third step of selecting a predetermined number of websites for each category among the websites extracted in the second step; And a fourth step of performing the evaluation on the websites selected through the third step, and providing the user with a website of the upper level for each category in the evaluation result along with the evaluation result.

4. Important uses of the invention

The present invention is used in information communication systems and the like.

Description

How to rate a website {Method For Evaluating A Web Site}

The present invention relates to a method for evaluating a website and a computer readable recording medium having recorded thereon a program for realizing the method.

With the development of the Internet, documents published on Internet-connected web servers are vast in number, and the information they provide is also diversified.

In this situation, computer users, who are called Internet navigators, are eager for a service that easily finds the information they want, and various search engines are appearing one after another according to their needs. That is, many search engines such as Yahoo, Simmani, AltaVista, Empas, etc. are attracting the attention of many users by highlighting their own unique search techniques.

These search engines have a keyword database to introduce each site for each registered site or a text search engine for each site. In other words, when a user enters a specific word or sentence that he / she is looking for, he or she uses keywords for an already built site or searches a whole sentence to search for a word or sentence that is similar to the word or sentence that the user inputs. Find and offer those sites.

However, while the conventional search engines such as the above find sites related to a certain topic on the Internet, the sites are often too many for the user to perform the first review, so users visit each site and There is a problem that it is very difficult to obtain the desired information.

In addition, despite the large number of sites resulting from the search of the conventional search engines and their quality, the search engines do not provide separate identification criteria for high quality sites. There is a problem that visits each site and brings the hassle of having to directly evaluate the site.

In addition, although the above-mentioned conventional search engines are evaluating the site to some extent, there is a problem that such evaluation is performed by people. In other words, due to the increase in the number of documents to be evaluated due to the increase of the website, even though it is difficult to provide satisfactory services to users unless the classification of the documents to be evaluated and the evaluation behavior itself are automated, this process is currently in progress. The problem is that it depends on human effort.

SUMMARY OF THE INVENTION The present invention has been made to solve the above problems, and to define the properties of high-quality sites and to automatically execute evaluation of each site on the Internet based on this definition, to provide users with high-quality sites among numerous sites. It is an object of the present invention to provide a computer-readable recording medium recording a site evaluation method and a program for realizing the method.

1 is a configuration diagram of an embodiment of an information communication system to which the present invention is applied.

2 is a flowchart of one embodiment of a website evaluation method according to the present invention;

3A and 3B illustrate one example of an evaluation questionnaire used in a website evaluation method according to the present invention.

According to an aspect of the present invention, there is provided a website evaluation method applied to an information communication system, comprising: a first step of classifying a category to be evaluated; A second step of automatically extracting websites to be used for evaluation and classifying the categories; A third step of selecting a predetermined number of websites for each category among the websites extracted in the second step; And a fourth step of performing the evaluation on the websites selected through the third step and providing a user with a website of a higher level for each category in the evaluation result along with the evaluation result.

The present invention also provides an information communication system including a high-capacity processor for website evaluation, comprising: a first function of classifying a category to be evaluated; A second function of automatically extracting websites to be used for evaluation and classifying them into the categories; A third function of selecting a predetermined number of websites for each category among the websites extracted by the second function; And a program for executing the evaluation of the websites selected through the third function, thereby realizing a fourth function of providing a user with a higher level website with the evaluation result for each category. A readable recording medium is provided.

In other words, the present invention is to maximize the automation of the current technology level of the items of the objective evaluation criteria of the objectively prepared in order to provide a method for effectively evaluating the various information content of the Internet website, and for the remaining items Experts in the field take the assessment and add up the results.

To do this, we first define some properties of high quality documents (referred to as "websites or web documents") (hereinafter simply referred to as "documents") and explain how to automatically evaluate how any document has these properties. .

First, the nature of high quality documents is explained.

Generally speaking, good quality documents have the following characteristics:

First, the contents of the document are organically related to each other.

Second, there is a connection between the appearance of each part of the document and the way in which the content of that part contributes to the content of the entire document.

Third, all the sentences that make up the content of the document are grammatical sentences.

Fourth, if a part of the document presupposes the existence of an object, then that object exists.

Due to the technical characteristics of the Internet operation and the social characteristics of the Internet use, the documents on the Internet may be considered as high quality documents in addition to the above properties.

That is, the time it takes for the document to be presented to the user should be short, and the amount of information provided by the document should be large.

Next, an automatic evaluation method for high quality documents will be described with reference to the drawings.

Hereinafter, a preferred embodiment according to the present invention will be described in detail with reference to FIGS. 1 to 3.

1 is a configuration diagram of an embodiment of an information communication system to which the present invention is applied.

As shown in the figure, the information communication system to which the present invention is applied operates servers 110 for providing various services, an evaluation server 130 for performing a website evaluation according to the present invention, and an evaluation server 130. It consists of an operator 138, a network 120, such as the Internet, which is in charge of communication between the servers 110 and the evaluation server 130, and the evaluator 150.

The evaluation server 130 is a web server 131 for communication with the network 120, a website extraction unit 133 for extracting website information from various information transmitted through the web server 131, extracted Evaluate each website using various website data of the analysis unit 134 for analyzing basic data from the information, the data management unit 135 for systematically managing the analyzed website data, and the data management unit 135. The evaluation unit 136 for controlling the evaluation data management unit 137 and all the components for providing the results to the user through the web server 131 while systematically managing the excellent website results of the evaluation of the data The control unit 132 for smooth transmission.

The server 110 providing various services is not only a general service providing server but also a search engine, a topic directory site, a web document award site, a server providing a specific recommendation-oriented site, and an individual service linked through the servers. Say server.

2 is a flowchart illustrating an embodiment of a website evaluation method according to the present invention, and FIGS. 3A and 3B are examples of evaluation questionnaires used in the website evaluation method according to the present invention.

Before conducting an assessment of the website, first establish a unique categorization scheme (level 3/4/5) (202). That is, for evaluating the website according to the present invention, the target websites are divided into 15 directories (topic areas) and managed by the data manager 135.

For example, Arts, Business, Computers, Games, Health, Home, News, Recreation, Reference, Geography (Regional), Science, Shopping, Society, Sports, World, etc.

This directory can be five stories deep by default, but a shallow subtopic may be three stories below the best topic.

That is, in the case of Computers, it is composed of five floors as follows.

Computers-Intenet-Email-Free-WebBased-ZZN Email (5F)

However, in the case of society, it is composed of three floors as follows.

Society-Issues-Older Citizens-Social Security Reform (3F)

The control unit 132 is the same or similar category from the existing search engine, subject directory site, web document award site, specific recommendation-oriented site, etc. 110 for each of the lowest level categories of the classification system through the process 202. It extracts the evaluation websites belonging to (if necessary, refer to the recommendation sites provided by the media or magazines) (204).

That is, it is a process of identifying the whereabouts of the documents (all of which have the same meaning as web documents or websites) to be evaluated. The reason why this process is important is, firstly, that it is impossible to extract documents to be evaluated randomly from a large number of web documents, considering the wide range of the objects. Second, documents that need to be distinguished first are already available to other documents. The assumption that is being referred to is valid.

The selection of the first evaluation target web document (204) utilizes existing Internet resources. This resource consists of four types:

First, there is a method using a server 110 that provides a search engine.

The majority of search engines take into account the degree of relevance when ordering documents to present search results. However, since this degree of relevance generally has nothing to do with the pros and cons of the document, the vast majority of search engines cannot play a significant role in the selection of documents for evaluation.

However, in the present invention, the unique ordering technique of the search engine 'google' is utilized. The engine decides whether to present first or later, depending on how many of the important sites in its database are mentioned. Thus, when searching by a subject, the first document presented is most frequently cited by other documents. That is, a search engine useful for extracting a website to be evaluated according to the present invention is selected from a number of search engines, and the first website to be evaluated is extracted from the search engine.

Here's how to get a list of the first sites to rate from "www.google.com":

For each leaf node in the Open Directory (the term that refers to the end node in the data structure tree) (hereafter simply called "leaf node"), do the following: At this time. The subtopic is called xyz, and the immediately preceding subtopic containing the subtopic is called abc.

First, use WiePost to get google's search results (WiePost http://www.google.com/search "? Q = xyz + abc & num = 30 & lc = www & btnG = Google + Search")

Next, the href (href parameter has a href = reference form) in this file, where reference is a destination address or destination file in URL format, creates one website element for each piece of information and adds it to site.xml. do. In other words, URL information of the necessary website is extracted and stored in the site.xml file.

Second, there is a method using the server 110 to provide a subject directory.

One of the few directories in the topic directory is 'magellan.excite.com'. The general directory has little room for use in the present invention because of the large number of documents included without special verification or schematic examination. For example, we don't use a directory like 'yahoo' because we can't target hundreds of thousands of documents provided by 'yahoo', and 'yahoo' doesn't give us any ranking material.

The following describes how to obtain a list of the first websites to be evaluated from 'magellan.excite.com'.

First, we use WieList program to collect directory information of 'magellan.excite.com'. For each leaf node xyz in this directory, determine the leaf node in the corresponding Open Directory. Next, WieCat http: //magellan.excite. com /.../ xyz ｜ Prune. The prune program extracts all the href information contained in the <TABLE> immediately after the <TABLE> containing the "Recommended Web Sites" and adds one website element from each href information in the sites.xml file. In other words, URL information of the necessary website is extracted and stored in the site.xml file.

Third, there is a method using a server 110 for providing a web document award site.

Web document award sites are generally awarded to about 100 sites per year, so it is natural to include all the documents awarded by them in the evaluation target site according to the present invention.

Fourth, there is a method using the server 110 to provide a specific recommendation-oriented site.

Many documents on the Internet refer to other documents on the Internet and usually provide links to them. All Internet document authors can explicitly include a list of recommended sites in their documents. The credibility of these referral sites may vary depending on the degree of knowledge or aesthetics of the referrer, but documents not included in the sources listed above may be recommended here. It can be used as a resource of the present invention. Among these, the so-called subject gateways, which introduce thousands of documents in a particular subject area, are important resources.

The following describes how to obtain a list of sites to be initially evaluated from the recommended site.

At this time, the recommendation sites (Subject Gateways) that have their own directory structure should use a method of obtaining a list of initial evaluation target sites from 'magellan.excite.com'. Selecting a primary evaluation target from a small-scale recommendation site that does not have its own directory structure (typically recommending dozens or hundreds of documents) is a bit more difficult. That is, the HTML document of the recommended site is decrypted to detect the subject area of the recommended site, and to determine which leaf node of the open directory corresponds to. This requires thesaurus.

For example, seed documents for obtaining a list of URLs of documents to be initially evaluated using the above four methods are as follows.

That is, www.google.com/magellan.excite.com/www.britannica.com/www.quoteland.com/www.webbyawards.com. In addition, it is possible to extract the evaluation target web documents through the other sites 110 satisfying the four methods, in addition to the above four methods can be extracted using another method that can extract the initial evaluation target.

The web documents selected first through the process 204 are stored in the website extraction unit 133 according to the classification system 202 of the data management unit 135.

The analysis unit 134 automatically analyzes and verifies the title of each website through the process 202 to determine whether the websites collected in each category are dealing with contents corresponding to the category (206).

According to the verification result, the reclassified websites are stored in the data manager 135 for each category and are managed (208).

Each category automatically selects the top 50 of the highest scoring websites, taking into account the frequency and ranking of referrals from existing recommendation / ranking sites (you can assign different weights based on the reliability / importance of the recommending websites). 210).

The evaluation unit 136 performs detailed and rigorous evaluation of the top 50 sites selected for each category by the website evaluation criteria consisting of 10 items (specific numbers may be modified arbitrarily).

That is, the website elements contained in the site.xml file through the evaluation target website extraction process 204 are all informed on the identity, recommendation, and classification elements when the above steps are completed. Will contain. However, these Web site elements lack information about evaluation. Thus, this evaluation process 212 is a process of specifying an evaluation for each website element.

The evaluation element has the following elements: access, structure, authority, quantity, novelty, comprehensiveness.

Of these, the incorporation of information into the comprehensiveness and novelty elements is done by human hands, but the filling of information into the rest of the elements is done by software. In order to save the evaluator's effort, only those documents that scored consistently in the criteria evaluated by the software can be considered as the evaluator's document.

The evaluation process of the connection speed among the items that can be automatically evaluated by the software is as follows.

Connection speed is the time it takes for the user to display the document in his browser. Factors affecting the connection speed are the clock speed of the central processing unit of the web server, the detection speed of the storage device, the capability of the communication line between the web server and the client, and the characteristics of the client. Among these factors, the communication line between the web server and the client is made up of several physically different lines, and the lines used to transfer actual information from the server to the client may vary from time to time.

This factor is the most important factor in determining the connection speed, and more than 95% of the time actually takes for the connection on the communication line.

The access speed of the document under evaluation shall be measured under the same conditions from any single client. Since the load on the communication line is different for each time zone, it is likely that a single measurement will yield distorted results. Therefore, the connection speed for each document is evaluated as the average of two measured values under the following conditions.

For example, measurements are made once between noon Saturday and 6:00 pm on a regular Saturday from a computer with an IP address of 168.188.82.xxx connected to the campus network of Chungnam National University and once between 9 am and 3 pm on Wednesday. The software used to measure is ping and the average of the two measurements is scored by the following scale.

That is, 5 points for 200 ms or less, 4 points for 201 ms to 400 ms, 3 points for 401 to 700 ms, 2 points for 701 ms to 1000 ms, and 1 point for 1001 ms or more are scored. If a "No answer host" message appears due to special circumstances, treat it specially.

It executes by constructing batch file from URL list of evaluation target documents. Since 1 hour is 3600 seconds, the connection speed with about 40000 URLs can be measured over 6 hours.

Obtain a file containing average measurements from a file containing Saturday afternoon measurements and a file containing Wednesday day measurements, and enter this file to evaluate the score in the access element of the evaluation element of each website element in sites.xml. Write and run a program that writes.

Next, a method of evaluating browser independence will be described.

Documents to be assessed will be penalized if they contain tags that are only recognized by certain browsers. That is, the evaluation score differs depending on how many tags such as bgsound, marquee, embed, noembed, multicol, spacer, nobr, wbr, nlink, xmp, listing, plaintext, keygen, layer, ilayer, and nolayer are included. When counting these, count the type.

5 points if it contains 0, 4 points if it contains 1-2, 3 points if it contains 3-5, 2 points, 9 or more if it contains 6-8 In the case of inclusion, it shall be one point. This process uses the HTML validation software.

Next, a method of evaluating link validity is as follows.

Other documents linked within the document must be accessible documents, and if the document contains links to inaccessible documents, the score is as low as that number.

That is, 5 points for non-accessible documents, 5 points for 1, 3 points for 2-3, 2 points for 4-6, 1 for 7 or more Let's do it. This process can be evaluated using a program called WieList.

Next, the method of evaluating HTML syntax is as follows.

Use the program that modified Neil Bowers' program weblint to check the HTML syntax of the evaluated document. As a result, the more errors, the lower the score. This check is based on the W3C HTML 3.2 and HTML 4.0 Recommendations and adopts the higher score.

5 points for 0 errors and 3 warnings, 4 points for 1-2 errors and 4-9 warnings, 3 points for 3-9 errors and 10-27 warnings. For example, 2 points for 10-30 errors and 28-81 warnings, and 1 point for 31 or more errors and 82 or more warnings.

This assessment is in fact redundant with the browser independence assessment. And it's going to be unnecessary by the end of three years, when the standardization of Web documents will turn into XML. By that time, HTML documents are already too old and will not even need to be evaluated.

That is, the evaluation unit 136 first automatically assigns an evaluation score to the one that can be automated as described above among a plurality of evaluation items, and manages the score for each website in the evaluation target data management unit 137.

And, for the items that a person must evaluate directly among the evaluation items, the relevant expert is requested to obtain evaluation scores for each item. That is, as shown in FIG. 3, the evaluation of the criterion [2] can be requested to the web design expert, and the criterion [8] and the criterion [10] to the experts of each field.

The evaluation process by the experts is as follows.

That is, by giving the ID and password to the evaluation member 150 belonging to the evaluation group to be able to access the evaluation server.

When the evaluator 150 connects to the web server 131 of the evaluation server 130 through the network 120, the control unit 132 determines which field the evaluator is evaluating by evaluating the data management unit ( 135) or the URL to be evaluated from the evaluation target data management unit 137 in a pull-down menu format to select one URL.

When one URL is selected, the site to be evaluated is displayed in the upper frame of the browser, and at the bottom of the frame, the URLs selected by the evaluator 150 and the criteria to be scored are presented in the form of radio buttons (FIGS. 3A and 3B). Reference).

When the evaluator 150 marks any one of the five grades and submits it to the web server 131, the control unit 132 weights the delivered score (the criterion [2] in FIG. 1 and the reference [8] and the reference [10] of FIG. 3B have a weight of 3), and store the result in a related record of the corresponding URL of the evaluation target data management unit 137.

Through the above process, the scores automatically evaluated for each website and the scores evaluated by experts are summed up in order of total scores to provide the top 30/40 sites by category (214). At this time, the top 30/40 sites provided by each category are explained in full and the method of efficient use according to the purpose is maximized, thereby maximizing the utilization value of consumers.

In addition, periodic evaluations are made on websites that have already been evaluated but whose contents have been changed and newly added websites, and the results are updated. It is checked whether the contents of the previously evaluated website have been changed by using a program, and the newly added website is periodically checked by repeating the above processes 204 to 210. In addition, the updated website and the added website are periodically reflected through the evaluation process 212 (214).

The present invention described above is capable of various substitutions, modifications, and changes without departing from the spirit of the present invention for those skilled in the art to which the present invention pertains, and the above-described embodiments and accompanying It is not limited to the drawing.

As described above, the present invention provides a user with automatic evaluation of the value of numerous websites so that users can visit the excellent websites and search for desired materials without additional time and effort. It works.

In other words, the present invention can provide a ranking by the automatic evaluation of the corresponding websites across the entire field of the category classification system, and can collect only the excellent sites in each category and provide them by the ranking, a portal that is practically helpful to users. It's a great effect that you can run your site.

Claims (7)

In the website evaluation method applied to the information communication system, A first step of classifying a category to be evaluated; A second step of automatically extracting websites to be used for evaluation and classifying the categories; A third step of selecting a predetermined number of websites for each category among the websites extracted in the second step; And A fourth step of performing the evaluation of the selected websites through the third step and providing a user with a higher level of the website by the evaluation result category together with the evaluation result; Website evaluation method comprising a. The method of claim 1, A fifth step of re-performing the evaluation periodically or aperiodically for the websites provided in the fourth step to provide an updated result to the user; Website evaluation method further comprising. The method according to claim 1 or 2, The process of automatically extracting the websites to be used as the evaluation targets of the second step comprises using at least one of a search engine, a topic directory site, a web document award site, and a specific recommendation center site. The method of claim 3, wherein The websites selected in the third step are websites that have high scores in consideration of the frequency and ranking of recommendations from the search engine, the subject directory site, the web document award site, and the specific recommendation center site. Way. The method according to claim 1 or 2, The fourth step, A sixth step of performing an evaluation on a portion corresponding to an automatic evaluation item for the websites selected through the third step; A seventh step of evaluating, by the experts, a portion that does not correspond to the automatic evaluation item for the website selected through the third step; And An eighth step of providing a user with a website of a higher level for each category of the evaluation result obtained by statistically evaluating the evaluation result performed in the sixth and seventh steps together with the evaluation result; Website evaluation method comprising a. The method of claim 5, The seventh step, A ninth step of providing an ID and a password to an evaluation member belonging to the evaluation group to access the evaluation server; A tenth step of determining, if the evaluator is accessing the evaluation server through a network, the evaluation member is evaluating which field the evaluator is; An eleventh step of presenting a website to be evaluated by the evaluator as a result of the evaluation and receiving the evaluation result; And A twelfth step of storing the evaluation result by synthesizing the input evaluation result Website evaluation method comprising a. In the information communication system with a large capacity processor for website evaluation, A first function of classifying a category to be evaluated; A second function of automatically extracting websites to be used for evaluation and classifying them into the categories; A third function of selecting a predetermined number of websites for each category among the websites extracted by the second function; And A fourth function of performing an evaluation on the websites selected through the third function and providing a user with a website of a higher level for each category of evaluation results together with the evaluation result; A computer-readable recording medium having recorded thereon a program for realizing this.