WO2006080627A1

WO2006080627A1 - A system and method for diagnosis, prescription and testing of english ability and a memory device recorded a program for diagnosis, prescription and testing of english ability

Info

Publication number: WO2006080627A1
Application number: PCT/KR2005/003129
Authority: WO
Inventors: Yong Myeong Kim
Original assignee: Yong Myeong Kim
Priority date: 2004-09-22
Filing date: 2005-09-21
Publication date: 2006-08-03
Also published as: KR100628983B1; KR20060027233A

Abstract

Disclosed are a system and a method for evaluating a user' s English ability based on accurate and objective criteria and providing the user with an individualized diagnosis and prescription to effectively improve the English ability and a method for evaluating a user's English ability and providing the user with an individualized diagnosis and prescription to improve the English ability using the same system. The English ability evaluation/diagnosis /prescription system having sub-systems including: an evaluation system for outputting evaluation data for English ability evaluation and evaluating a user' s English ability based on answer data inputted by the user; a sentence development stage classifier system for classifying sentences included in the answer data into a pertinent development stage and transferring results of classification to the evaluation system; a diagnosis/prescription system for diagnosing the user's English ability based on results of evaluation outputted from the evaluation system and providing the user with an individualized prescription; a data storage unit for storing data necessary for the English ability evaluation, diagnosis and prescription; and a control unit for controlling overall operations of the subsystems to provide accurate evaluation, diagnosis and prescription concerning the user's English ability.

Description

A SYSTEM AND METHOD FOR DIAGNOSIS, PRESCRIPTION AND TESTING OF ENGLISH ABILITY AND A MEMORY DEVICE RECORDED A PROGRAM FOR

DIAGNOSIS, PRESCRIPTION AND TESTING OF ENGLISH ABILITY

BACKGROUND OF THE INVENTION

Field of the invention

The present invention relates to an English ability evaluation/diagnosis/prescription system, and more particularly to an English ability evaluation/diagnosis/prescription system for accurately evaluating a user' s English ability with an objective and reliable method and offering an individualized diagnosis and prescription for how to improve one' s English ability, a method for evaluating English ability and offering individualized diagnosis and prescription using the same system and a computer-readable memory device for recording English ability evaluation, diagnosis and prescription programs .

Description of the Prior Art

With the rapid globalization, the demand for foreign language skills has increased and the education industry offering a variety of language training institutes and programs has also grown significantly over the past decades . Particularly, proficiency in English is highly demanded as English has become the universal language of our time . Non- English-speaking countries encourage educational environments and programs for teaching English to their citizens from very young ages . In Korea, education systems have been changed to teach English as a required subject from elementary school to meet the growing need for English education in this globalized world. Some parents have an intense fervor for English education and are eager to make their children learn English from early childhood or even infancy. With the increasing use of the Internet, online language education has been greatly activated with no less importance or effect than offline education. Language learners generally use small-sized cassette recorders or MP3 players to listen and practice foreign languages . With the desire to improve English language skills, many of those learners study English at language institutes offering classes from the early morning to late in the evening. However, most teaching/learning methods and programs used in English education cannot be useful if they are not offered based on the accurate evaluation and diagnosis of each learner' s English language skills . It is not possible to learn English effectively through the education offered without a comprehensive evaluation of the learners or through the simply repeated study using specific textbooks . Although a variety of tests such as TOEFL or TOEIC are available to assess proficiency in English, those tests also have some minor problems . Test applicants should pay application fees in US currency and take the tests such as TOEFL or TOEIC for several hours without rest . In addition, scores obtained on such tests cannot be considered completely reliable criteria that ensure a comprehensive evaluation of each applicant' s English proficiency level .

As stated above, a variety of English education methods and programs have been suggested to date . However, those methods and programs are not much helpful if they neither provide an obj ective and accurate evaluation of English skills nor suggest an individualized diagnosis and prescription to improve each learner' s English skills .

SUlMMARY OF THE INVENTION

Accordingly, the present invention has been made to solve the above-mentioned problems occurring in the prior art, and an object of the present invention is to provide an English ability evaluation/diagnosis/prescription system for evaluating a user' s English ability based on accurate and obj ective criteria and providing the user with an individualized diagnosis and prescription to effectively improve the English ability, and a method for evaluating a user' s English ability and providing the user with an individualized diagnosis and prescription to improve the English ability using the same system.

In order to accomplish this obj ect, there is provided an English ability evaluation/diagnosis/prescription system having sub-systems including : an evaluation system for outputting evaluation data for English ability evaluation and evaluating a user' s English ability based on answer data inputted by the user; a sentence development stage classifier system for classifying sentences included in the answer data into a pertinent development stage and transferring results of classification to the evaluation system; a diagnosis/prescription system for diagnosing the user' s English ability based on results of evaluation outputted from the evaluation system and providing the user with an individualized prescription; a data storage unit for storing data necessary for the English ability evaluation, diagnosis and prescription; and a control unit for controlling overall operations of the sub-systems to provide accurate evaluation, diagnosis and prescription concerning the user' s English ability.

The sentence development stage classifier system analyzes the structural complexity and difficulty level of the sentences in the answer data inputted by the user in response to the evaluation data and classifies the sentences in the answer data into a pertinent development stage based on the analysis .

The sentence development stage classifier system includes : a constituent movement detector for classifying the answer data by sentence pattern (simple, semi-complex or complex) , detecting any grammatical structure change in each sentence of the answer data as compared to a standard grammatical structure of each sentence pattern and outputting a first variable value ( α ) corresponding to the detected grammatical structure change; a word order change detector for detecting any change of word order in each sentence of the answer data as compared to a standard word order in each sentence pattern and outputting a second variable value ( β ) corresponding to the detected word order change; and a sentence combination detector for detecting any grammatical structure change in an embedded clause of a complex sentence included in the answer data and outputting a third variable value ( Y ) corresponding to the detected grammatical structure change .

The diagnosis/prescription system includes a learning system for outputting evaluation data-to-be-learned which corresponds to the evaluation results received from the evaluation system, evaluating answer data inputted in response to the evaluation data-to-be-learned and adjusting the difficulty level of the data-to-be-learned to allow the user to keep learning effectively.

The evaluation system includes : a storage unit for storing a value inputted from the sentence development stage classifier system which corresponds to the development stage of each sentence; an extractor/output unit for extracting the development stage value stored in the storage unit and assessing the extracted value to produce a correct answering rate in the answer data inputted by the user in response to the evaluation data of a specific stage; and a determination unit for determining the user' s English proficiency level based on the correct answering rate .

The evaluation system further includes : a text output means for outputting the evaluation data in text; a speech output means for outputting the answer data in speech; a speech input means for inputting the answer data in speech; and a text input means for inputting the answer data in text .

The learning system outputs the evaluation data-to-be- learned to the evaluation system, evaluates the answer data inputted in response to the evaluation data-to-be-learned based on a value received from the sentence development stage classifier system which corresponds to the development stage of the answer data and determines whether to keep providing data-to- be-learned or adjust the difficulty level of the data-to-be- learned.

The data storage unit consists of : a first data storage unit for storing evaluation data in text; a second data storage unit for storing evaluation data in speech; a third data storage unit for storing diagnosis/prescription data corresponding to each stage; and a fourth data storage unit for storing reference rates of correct answering for passing the respective stages of evaluation.

The evaluation data includes sentence construction ability evaluation data, translation ability evaluation data and dialogue ability evaluation data .

In accordance with another aspect of the present invention, there is provided a method for evaluating a user' s English ability and providing the user with an individualized diagnosis and prescription to improve the user' s English ability, which comprises the steps of : outputting evaluation data classified by stage according structural complexity and difficulty levels of English sentences; analyzing the structural complexity and difficulty level of sentences in answer data inputted by the user in response to the evaluation data to classify each sentence in the answer data into a pertinent development stage; calculating a correct answering rate in the answer data and evaluating the user' s English ability based on the correct answering rate; and providing the user with an individualized diagnosis and prescription based on results of evaluation.

The evaluation data may be outputted sequentially by stages classified according to the structural complexity and difficulty levels of sentences or randomly regardless of the stages . The evaluation data consists of example sentences and questions outputted in text or speech . The example sentences and corresponding questions are sequentially outputted in predetermined time intervals . Preferably, each example sentence and question should be displayed in a flash form that disappears after a few seconds . The answer data can be inputted in text or speech.

The step of classifying each sentence in the answer data into a pertinent development stage includes : recognizing a pattern of the sentence; detecting any change of grammatical structure and word order in the sentence in view of standard grammatical structure and word order of the pattern; and classifying the sentence into a pertinent development stage based on variable values assigned according to the detected change in grammatical structure and word order . The step of evaluating the user' s English ability includes : storing a data value corresponding to the answer data of each stage; extracting the stored data value of each stage; evaluating the extracted data value to calculate a correct answering rate in the answer data; comparing the calculated correct answering rate with a reference correct answering rate previously set for passing the stage; and determining a corresponding English ability level of the user who inputted the answer data .

The English ability level of the user is determined upon comparison between the correct answering rate in the answer data inputted in response to the evaluation data by stage with the reference rate previously set for passing each stage .

The method for evaluating a user' s English ability and providing the user with an individualized diagnosis and prescription further comprises the step of: providing evaluation data-to-be-learned which corresponds to the results of diagnosis and prescription to enable the user to keep learning using evaluation data suitable to the user' s English ability level . The learning step includes : outputting evaluation data-to- be-learned; allowing the user to input answer data in response to the evaluation data-to-be-learned; transferring the answer data to a sentence development stage classifier system; receiving a data value corresponding to the development stage of each sentence in the answer data from the sentence development stage classifier system; storing the data value corresponding to the development stage of each sentence in the answer data; evaluating the data value to calculate a correct answering rate in the answer data; comparing the calculated correct answering rate with a reference correct answering rate previously set for the stage; and determining whether the user' s English ability level has changed.

In accordance with still another aspect of the present invention, there is provided a computer-readable memory device for recording programs for English ability evaluation, diagnosis and prescription. The programs are used to provide a user with evaluation data in text or speech which is classified by development stage according to the structural complexity and difficulty levels of English sentences, receive answer data inputted by the user in text or speech in response to the evaluation data, analyze the structural complexity and difficulty level of each sentence in the answer data to classify the sentence into a pertinent development stage, evaluate the user' s English ability based on a correct answering rate in the answer data, provide the user with an individualized diagnosis and prescription and, if necessary, output evaluation data-to-be- learned so that the user can keep learning to improve his or her English ability.

Hereinafter, preferred embodiments of an English ability evaluation/diagnosis/prescription system, a method for evaluating English ability and offering an individualized diagnosis and prescription using the same system and a computer-readable memory device for recording English ability evaluation, diagnosis and prescription programs will be explained in detail .

The above system and method analyze the structural complexity and difficulty level of English sentences inputted by the user in answer data to accurately classify the sentences into a pertinent development stage, provide the user with objective accurate diagnosis and prescription concerning the user' s English ability based on the classified development stage to effectively improve the user' s English ability through individualized instruction and learning methods . . The above system and method quantify mechanisms for constituent movement and word order change in English sentences and formulate the development stages of the English sentences using sentence complexity functions . The system detects any movement of constituent or change of word order in each sentence and applies the detected movement of constituent or change of word order to a sentence complexity function in order to accurately classify the sentence into a pertinent development stage using the sentence complexity function and thereby accurately evaluate the user' s present level of English proficiency. The system makes a diagnosis of the user' s current English ability based on the classification of the development stage of the user' s English sentences and thereby provides the user with an effective and systematic prescription and instruction as to what and how the user should learn.

A variety of tests currently available to assess proficiency in English have limitations in providing accurate evaluation, diagnosis and prescription concering English ability. Test applicants should pay application fees and take the tests for several hours without rest . Scores obtained on such tests cannot be considered completely reliable criteria for a comprehensive evaluation of each applicant' s English proficiency level . The system and method according to the present invention can accurately and easily evaluate a user' s English ability at a low cost and provide the user with an individualized diagnosis and prescription.

BRIEF DESCRIPTION OF THE DRAWINGS

The above and other objects, features and advantages of the present invention will be more apparent from the following detailed description taken in conjunction with the accompanying drawings, in which:

FIG. 1 is a block diagram of an English ability evaluation/diagnosis/prescription system according to the present invention;

FIG. 2 is a block diagram of a sentence development stage classifier system included in an English ability evaluation/diagnosis/prescription system according to the present invention; FIG. 3 is a block diagram of an evaluation system included in an English ability evaluation/diagnosis/prescription system according to the present invention;

FIG. 4 is a view showing evaluation data including an example sentence and a question displayed on a text output means of an evaluation system during an English sentence construction test according to the present invention;

FIG. 5 is a view showing evaluation data including an example sentence and a question displayed on a text output means of an evaluation system during an English sentence translation test according to the present invention;

FIG. 6 is a block diagram of a diagnosis/prescription system included in an English ability evaluation/diagnosis /prescription system according to the present invention;

FIG. 7 is a block diagram of a learning system included in an English ability evaluation/diagnosis/prescription system according to the present invention;

FIG. 8 is a block diagram of a data storage unit included in an English ability evaluation/diagnosis/prescription system according to the present invention;

FIGs . 9 through 19 are views showing examples of evaluation data in different stages which are provided in text according to a preferred embodiment of the present invention; FIG. 20 is a flow chart showing a process for evaluating English ability and offering an individualized diagnosis and prescription according to the present invention;

FIG. 21 is a flow chart showing a process for classifying sentence development stages according to the present invention; FIG. 22 is a flow chart showing a process for evaluating English ability according to the present invention;

FIG. 23 is a flow chart showing a process for learning English according to the present invention; and

FIGs . 24 and 25 are tables showing test results obtained using an English ability evaluation/diagnosis/prescription system and method according to the present invention.

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS

Hereinafter, a preferred embodiment of the present invention will be described with reference to the accompanying drawings . In the following description and drawings, ' the same reference numerals are used to designate the same or similar components, and repetition of the description on the same or similar components will be omitted.

FIG. 1 is a block diagram of an English ability evaluation/diagnosis/prescription' system according to the present invention. The system includes sub-systems, such as a sentence development stage classifier system 100 , an evaluation system 200 and a diagnosis/prescription system 300, a data storage unit 400 for storing various data necessary for the operations of the system and a control unit 500 for controlling the overall operations of the system.

The sentence development stage classifier system 100 analyzes the structural complexity and difficulty level of sentences included in answer data which has been inputted by a user in response to evaluation data and determines development stages of the sentences in the answer data based on results of analysis . This system is disclosed in detail in an earlier patent application (Korean Patent Application No . 10-2004-64089") filed by the same applicant, entitled "System and Method for English Development Stage Classifier system and Computer-Readable Memory Device for Recording English Development stage Classification Program. "

FIG . 2 is a block diagram of the sentence development stage classifier system included in the English ability evaluation/diagnosis/prescription system according to the present invention. The sentence development stage classifier system 100 includes a sentence recognizer 110, a pattern recognizer 120, a constituent movement detector 130, a word order change detector 140, a sentence combination detector 150 and a stage classifier 160. The sentence recognizer 110 has a program for recognizing answer data (English sentences) inputted in text or speech through the evaluation system 200. The pattern recognizer 120 has a program for recognizing a pattern of each sentence included in the answer data . The pattern recognizer 120 may use a recognition technique generally used in conventional automatic translators . In other words, automatic translators currently available to translate English into Korean or Korean into English basically recognize patterns of sentences in original text, select Korean or English words corresponding to the words in the original text and combine the selected words according to the recognized patterns . Programs used in such automatic translators can classify sentence patterns into simple sentences, semi- complex sentences and complex sentences .

The pattern recognizer 120 recognizes the pattern of each sentence which can be simple, semi-complex or complex. Sentences in any of the following patterns are recognized as simple sentences : [VP] (ADJ) / [VC] (ADJ) / [VO] (ADJ) / [SV] (ADJ) / [SVP] (ADJ) /

[SVC] (ADJ) / [SVO] (ADJ) / [SVPO] (ADJ)

Sentences in any of the following patterns are recognized as semi-complex sentences :

[SVOC] (ADJ) / [SVOO] (ADJ) / [SVOPO] (ADJ) Lastly, sentences in any of the following patterns are recognized as complex sentences :

[ [SVO] + [SVOC] ] (ADJ) / [ [SVOC] + [SVOC] ] (ADJ)

The pattern recognizer 120 further includes a database (not shown) that stores information about variables in each of the above patterns in order to determine the pattern of any inputted sentence (simple, semi-complex, complex or multiple-complex) .

The sentence development stage classifier system 100 classifies the sentence development stages of simple, semi- complex, complex and multiple-complex sentences using three English sentence construction mechanisms ( Y SCM/ α CMM/J3 BOM) . The sentence development stage classifier system 100 includes a constituent movement detector 130 for determining movement (or addition/insertion) of any constituent of an English sentence based on standard English sentence structures, a word order change detector 140 for determining any change of word order based on a standard word order in each sentence pattern and a sentence combination detector 150 for determining whether a constituent of an embedded clause in a complex sentence is moved within or out from the embedded clause .

The constituent movement detector 130 outputs a variable value ( α : a first variable value) according to the grammatical structure change in an inputted sentence as compared to the standard grammatical structure of each sentence pattern. The word order change detector 140 outputs a variable value ( β : a second variable value) according to the word order change in the inputted sentence as compared to the standard word order of each sentence pattern. The sentence combination detector 150 outputs a variable value ( y : a third variable value) upon determination whether a constituent of an embedded clause is moved within or out from the embedded clause . The sentence development stage classifier system 100 further includes a stage classifier 160 for classifying the development stages of sentences in each pattern according to the first to third variable values outputted from the constituent movement detector 130, word order change detector 140 and sentence combination detector 150, generating a value corresponding to each development stage and outputting the generated value to the evaluation system 200.

The sentence development stage classifier system 100 further includes a database (not shown) for storing inputted sentences that cannot be classified by development stage. Sentences that cannot be classified by development stage refer to those to which no development stage can be assigned, for example, completely wild sentences and culture-specific sentences used under no standard guideline for word order or movement of sentence constituents . The sentence development stage classifier system 100 classifies the development stages of sentences included in the inputted answer data by pattern (simple, semi-complex or complex) . For simple sentences included in the answer data, the sentence development stage classifier system 100 classifies the sentence development stages depending on variables α and β in a simple sentence complexity function f{a , β ) = [ α CMM/β BOM] . For example, if a sentence in the answer data inputted by the user has neither a movement of constituent nor a change of word order as compared to the standard simple sentence pattern (SVO pattern) , the sentence in the answer data will be classified into the lowest stage IX of possible development stages of simple sentences . If an SVO sentence has no change in grammar or word order despite a constituent moved or added to the beginning of the sentence, the sentence will be classified into the second lowest stage 2X, i . e . , one stage higher than IX. If an SVO sentence in the answer data has a change in word order due to a new constituent added to the sentence or a constituent moved to the beginning from the middle of the sentence, the sentence will be classified into 3X, i . e . , one stage higher than 2X. If a constituent is moved to the beginning of an SVO sentence and another constituent is moved within the sentence, thereby causing a change in word order, the sentence will be classified into the highest stage 4X of the possible development stages of simple sentences .

For semi-complex sentences included in the answer data, the sentence development stage classifier system 100 classifies the sentences into development stages 5Xa, 5Xb, 5Xc and 5Xd. The sentence development stage classifier system 100 obtains four types of [SCM ( [CMM] / [BOM] ) ] combinations from hierarchical combinations of variables α and β in a semi-complex sentence complexity function f(α , β ) = [SCM ( [ α CMM/13 BOM] ) ] , quantifies the structural complexity of the semi-complex sentences based on the combination values and thereby classifies the semi-complex sentences into the four development stages 5Xa, 5Xb, 5Xc and 5Xd. The same scale values of [ α CMM/β BOM] of ECM at stages IX, 2X, 3X and 4X are applied to [ α CMM/β BOM] at stages 5Xa, 5Xb, 5Xc and 5Xd. In other words, the mechanism for classifying simple sentences into the development stages IX, 2X, 3X and 4X is used in the mechanism for classifying semi-complex sentences into the development stages 5Xa, 5Xb, 5Xc and 5Xd. It is possible to explain how semi-complex sentences are classified into stages 5Xa, 5Xb, 5Xc and 5Xd based on the mechanism of classifying simple sentences into stages IX, 2X, 3X and 4X. Semi-complex sentences and simple sentences constitute an implicational scale.

For complex sentences each composed of a main clause and an embedded clause, the sentence development stage classifier system 100 classifies the complex sentences using the simple sentence complexity function f^"(α , jβ ) = [ α CMM/Jβ BOM] (when the main clause is a simple sentence) or the semi-complex sentence complexity function f(a , β ) = [SCM ( [ α CMM/β BOM] ) ] (when the main clause is a semi-complex sentence) . The sentence development stage classifier system 100 obtains five types of [SCM ( [CMM] / [BOM] ) ] combination from hierarchical combinations of variables α , β and Y in an embedded clause complexity function f(α , β , Y ) = [Y SCM ( [ α CMM/Jβ BOM] ) ] , quantifies the structural complexity of embedded clauses in English based on the five ECM combination values and thereby classifies the embedded clauses into the five development stages 1X2, 2X2, 3X2, 4X2 and 5X2. The same scale values of [ α CMM/β BOM] of ECM at stages IX, 2X, 3X and 4X are applied to [ α CMM/β BOM] at stages 1X2, 2X2, 3X2, 4X2 and 5X2. In other words, the mechanism for classifying simple sentences into the development stages IX, 2X, 3X and 4X is used in the mechanism for classifying embedded clauses into the development stages 1X2, 2X2, 3X2, 4X2 and 5X2. It is possible to explain how semi-complex sentences are classified into stages 5Xa, 5Xb, 5Xc and 5Xd, or how embedded clauses are classified into stages 1X2, 2X2, 3X2, 4X2 and 5X2 based on the mechanism of classifying simple sentences into stages IX, 2X, 3X and 4X. Embedded clauses, semi-complex sentences and simple sentences constitute an implicational scale . Unlike semi-complex or simple sentences, embedded clauses has one more development stage 5X2. A complex sentence with a constituent moved out from an embedded clause during the combination of two simple sentences, i . e . , a main clause and the embedded clause, is classified into the fifth development stage 5X2. When the structural complexity of an embedded clause is formulated in such a way, only the movement ( [ ±CMM] ) of a critical factor associated directly with the embedded clause (a connective, conjunction, interrogative, relatival or the like) and the change in word order ( [ +BOM] ) will be considered to apply the ECM mechanism of the embedded clause and obtain variables α , β and Y in [Y SCM ( [ α CMM/β BOM] ) ] .

Therefore, the structure of the embedded clause itself, for example, whether the embedded clause is an SVO pattern or an

SVOOC pattern, will not be further considered to formulate the complexity of the embedded clause . According to the implicational scale, a learner who knows the structure of embedded clauses is deemed to have already learned the structures of simple and semi-complex sentences .

To determine the development stage of a complex sentence which is composed of a main clause and an embedded clause, the development stage of the embedded clause is first formulated using the simple sentence complexity function and the semi- complex sentence complexity function and then the development stages of the main clause and the embedded clause are combined together. Therefore, the development stage of a complex sentence can be represented by "KX2 + kx. " For example, "3X2 + 2x" stage of a complex sentence implies that the main clause and embedded clause of the complex sentence are of development stages 2x and 3X2, respectively.

As explained above, the sentence development stage classifier system 100 recognizes the pattern (simple, semi- complex or complex) of sentences included in the answer data. When the sentences are recognized as simple pattern sentences, the system 100 classifies the sentences into development stages IX, 2X, 3X and 4X depending on variables α and β in the simple sentence complexity function _f(α , β ) = [ α CMM/β BOM] . When the sentences in the answer data are semi-complex, the system 100 obtains four types of [SCM ( [CMM] / [BOM] ) ] combinations from hierarchical combinations of variables α and β in the semi- complex sentence complexity function _f(α , β ) = [SCM ( [ α CMM /β BOM] ) ] , quantifies the structural complexity of the semi- complex sentences based on the four combination values and thereby classifies the semi-complex sentences into four development stages 5Xa, 5Xb, 5Xc and 5Xd. When the sentences in the answer data are in a complex pattern composed of a main clause and an embedded clause, the system 100 formulates the development stage of a main clause in each complex sentence using the simple sentence complexity function or the semi-complex sentence complexity function and classifies an embedded clause in each sentence into development stage 1X2, 2X2, 3X2, 4X2 or 5X2 depending on variables α , β and Y in the embedded clause complexity function f(a , β , Y ) = [Y SCM ( [ α CMM/β BOM] ) ] . Then the development stages of the main clause and the embedded clause of each sentence are combined together to determine the overall development stage of the complex sentence . Moreover, the sentence development stage classifier system 100 can classify the development stages of multiple-complex sentences, each of which is composed of two or more embedded clauses . The embedded clause complexity function as explained above is used to classify the development stages of multiple- complex sentences .

In the English ability evaluation/diagnosis/prescription system according to the present invention, the evaluation system 200 includes a storage unit 210, an extractor 220, an output unit 230, a comparator 240 and a determination unit 250 as illustrated in FIG. 3. The storage unit 210 stores data values inputted from the stage classifier 160 of the sentence development stage classifier system 100 which correspond to the development stages (from IX to X3) of answer data. The extractor 220 extracts the stored data values in different stages . The output unit 230 assesses the extracted data values in different stages to produce a correct answering rate. If a user inputs answers to the questions offered as evaluation data in a specific stage, the correct answering rate can tell how many answers inputted by the user are correct . Upon output of a correct answering rate, the comparator 240 compares the outputted rate with a preset reference rate (%) at or above which users should obtain in answer data to pass a specific stage of English ability evaluation. The same reference rate (for example, 100%) of correct answering may be set for every stage of evaluation data . Alternatively, different reference rates may be set for different stages of evaluation data. For example, 100% can be set as the reference correct answering rate for passing stage IX of evaluation (lowest development stage of simple sentences) , 70% for passing stage 5Xc of evaluation (middle development stage of semi-complex sentences) and 50% for passing stage 5X2 of evaluation (highest development stage of complex sentences) . Based on the results of comparison between the correct answering rate in the answer data and the reference rate in the pertinent stage, the determination unit 250 finally determines the English ability level of the user who inputted the answer data . If the user has obtained higher rates of right answers than the reference rates preset for stages IX to 4X of simple sentences but a lower rate than the reference rate preset for stage 5Xa of semi-complex sentences, the determination unit 50 will determine that the user' s English ability is stage 4X. The evaluation system 200 further includes a text output means 260 for outputting the evaluation data in visible text, a speech output means 270 for outputting the answer data in audible speech, a speech input means 280 for inputting the user' s answer data in speech and a text input means 290 for inputting the user' s answer data in text . The answer data inputted through the speech input means 280 and the text input means 290 is provided to the sentence development stage classifier system 100.

The evaluation data outputted from the text output means 260 and the speech output means 270 includes questions relating to various sentence patterns (such as simple, semi-complex and complex) classified in various stages from lowest to highest according to the structural complexity and difficulty level . Structurally, the evaluation data consists of question data and example sentence data. In regards to content, the evaluation data consists of sentence construction ability evaluation data, translation ability evaluation data and dialogue ability evaluation data.

The text output means 260 outputs the evaluation data for assessing the user' s English ability in text, including example sentences and questions, to a display unit . Also, the text output means 260 outputs individualized diagnosis and prescription data generated based on the results of evaluation to the display unit to be seen by the user. Preferably, the display unit should be a display device such as a computer monitor and should be connected to a printer. The text output means 260 outputs the evaluation data (text data) stored in the data storage unit 400 and the answer data (text data) inputted by the user to the display unit under the control of a data driver (not shown) controlled by the control unit 500. The text output means 260 also outputs evaluation results, diagnosis data generated based on the evaluation results and prescription data generated based on the diagnosis data to the display unit . Preferably, example sentences and questions outputted to the display unit should disappear after a few seconds in a flash-like fashion.

The speech output means 270 outputs various evaluation data in speech for assessing a user' s English ability. Preferably, the speech output means 270 should be a speaker that outputs evaluation data formed in speech, including example sentences and questions, to be heard by the user under the control of the control unit 500.

The speech input means 280 inputs speech produced by the user to answer questions in the evaluation data. Preferably, the speech input means 280 should be a microphone. The speech input means 280 is used for the user to input spoken answer data in response to the example sentences and questions outputted through the text output means 260 or the speech output means 270. The user' s speech inputted through the speech input means 280 is transferred to the sentence development stage classifier system 100.

The text input means 290 used to input various answer data for the English ability evaluation should preferably be a keyboard, a tablet or the like. Users can input answer data in text in response to the examples sentences and questions outputted in speech or text . The answer data inputted through the text input means 290 is transferred to the sentence development stage classifier system 100. Also, users can select a desired test for English ability evaluation through the text input means 290. Three tests, i . e . , SCT (English Sentence Construction Test) , STT (English Sentence Translation Test) and DCT (English Dialogue Test) , are available to evaluate English ability.

When SCT is selected to evaluate a user' s ability to construct English sentences, example English sentences and questions (for example, "Change to an interrogative sentence" and "Combine two sentences") are outputted in text or speech through the text output means 260 or the speech output means 270. When the user inputs answer data to the speech input means 280 or the text input means 290, the sentence development stage classifier system 100 classifies the sentence development stage of the inputted answer data to assess the user' s ability to construct English sentences . As shown in Table 1, SCT has the following four test types according to the example sentences/questions output type and the answer data input type : [Table 1]

For example, a test type that outputs example sentences and questions in text and allows a user to input answers in text may be used to classify the stage of the answer data and thereby assess the user' s ability to construct English sentences . At this time, text display on the text output means 260 is as illustrated in FIG. 4.

STT is a test for assessing a user' s ability to translate Korean sentences into English. This test has two types of evaluation data output; one for outputting evaluation data in speech through the speech output means 270 and the other for outputting evaluation data in text through the text output means 260. Also, this test has two types of answer data input; one for inputting answer data in speech through the speech input means 280 and the other for inputting answer data in text through the text input means 290. For example, an STT test type that outputs example sentences and questions through the text output means 260 and allows the user to input answers in text through the text input means 290 may be used to assess the user' s translation ability. At this time, text displayable on the text output means 290 is as illustrated in FIG. 5. Preferably, example sentences and questions should be displayed in a flash form that disappears after a few seconds .

DCT is a test for assessing a user' s ability to communicate effectively in English. This test presents dialogue sentences through the text output means 260 or the speech output means 270, with certain phrases or sentences blanked out . The user has to speak or write answers to be inserted in blanks through the speech input means 280 or the text input means 290.

As illustrated in FIG. 6, the diagnosis/prescription system 300 of the English ability evaluation/diagnosis /prescription system according to the present invention includes a diagnosis data extractor 310 and a prescription data extractor 320. The diagnosis/prescription system 300 diagnoses current English language skills of a user based on the results of test performed using the evaluation system 200 and provides the user with an individualized prescription and instruction to improve his or her English ability. To be specific, the diagnosis data extractor 310 receives evaluation result data inputted from the evaluation system 200, extracts corresponding diagnosis data from the data storage unit 400 and outputs the extracted diagnosis data through the text output means 260 or the speech output means 270. The prescription data extractor 320 extracts prescription data corresponding to the evaluation result data from the data storage unit 400 and outputs the extracted prescription data through the text output means 260 or the speech output means 270.

The diagnosis/prescription system 300 further includes a learning system 300a for outputting evaluation data for learning according to the evaluation results received from the evaluation system 200 and evaluating answer data inputted in response to the evaluation data for learning. The learning system 300a outputs the evaluation data for learning to the evaluation system 200 and evaluates the answer data inputted in response to the evaluation data for learning based on a value corresponding to the development stage of the answer data classified by the sentence development stage classifier system 100. Then the learning system 300a determines whether to keep providing learning data or adjust the difficulty level of the learning data.

As illustrated in FIG. 7, the learning system 300a includes a learning stage selector 300b, an evaluation data-to-be-learned extractor 300c, a storage unit 30Od, an extractor 30Oe, an output unit 30Of, a comparator 30Og and a stage adjuster 30Oh. The learning stage selector 300a generates a window for selecting a desired language learning stage and outputs the generate window through the text output means 260. The evaluation data-to-be- learned extractor 300c extracts evaluation data for learning which corresponds to the selected language learning stage and outputs the extracted evaluation data through the text output means 260 or the speech output means 270. The storage unit 30Od, extractor 30Oe, output unit 30Of and comparator 30Og do the same function as the storage unit 210, extractor 220, output unit 230 and comparator 240 included in the evaluation system 200. Accordingly, no further explanation will be made on those units . The comparator 30Og compares the correct answering rate that the user obtained in the evaluation data for learning with a reference correct answering rate that has been set for the selected language learning stage. When the obtained correct answering rate is higher than the reference rate, the stage adjuster 30Oh increases the stage of the user' s English ability by one and informs the evaluation data-to-be-learned extractor 300c of the increase of stage . Then the evaluation data-to-be- learned extractor 300c extracts evaluation data for learning which corresponds to the increased stage and outputs the extracted evaluation data through the text output means 260 or the speech output means 270, thereby keeping providing the user with English learning data .

The data storage unit 400 of the English ability evaluation/diagnosis/prescription system according to the present invention stores evaluation data such as example sentences and questions for assessing a user' s English ability, diagnosis/prescription data generated according to the results of evaluation and learning data for the user' s English learning.

FIG. 8 is a block diagram of the data storage unit 400 of the English ability evaluation/diagnosis/prescription system according to the present invention. The data storage unit 400 consists of a first data storage unit 400a for storing evaluation data in text, such as example sentences and questions, in different stages (from IX to X3) for assessing English ability, a second data storage unit 400b for storing evaluation data in speech, such as example sentences and questions, in different stages, a third data storage unit 400c for storing diagnosis/prescription data corresponding to each stage and a fourth data storage unit 40Od for storing reference rates of correct answering for passing the respective stages of evaluation . The first and second data storage units 400a and 400b store evaluation data for learning as well as the evaluation data in text or speech.

The first data storage unit 400a stores a plurality of example sentences in each stage. The more example sentences are stored, the more accurate the evaluation can be made on English ability. FIGs . 9 through 19 are views showing examples of evaluation data in different stages which are provided in text according to a preferred embodiment of the present invention.

The overall operations of the English ability evaluation /diagnosis/prescription system according to the present invention are controlled by the control unit 500. The control unit 500 interconnects the sub-systems of the English ability evaluation/diagnosis/prescription system (i . e . , the sentence development stage classifier system 100, evaluation system 200 and diagnosis/prescription system 300) with each another to accurately and objectively assess a user' s English ability and thereby controls all the sub-systems to provide the user with an individualized diagnosis and prescription to improve his or her English ability.

Under the control of the control unit 500 of the English ability evaluation/diagnosis/prescription system, evaluation data stored in text or speech in the data storage unit 400 is outputted through the text output means 260 or the speech output means 270. When the user inputs answer data, the control unit 500 transfers the answer data to the sentence development stage classifier system 100 in which the development stage of the answer data will be classified. The user' s English ability is evaluated according to the classified development stage of the answer data. Based on the results of evaluation, individualized diagnosis and prescription are provided to the user. If the user wishes, he or she can keep learning English using evaluation data provided for learning purpose .

Hereinafter, a process for evaluating a user' s English ability and providing the user with an individualized diagnosis and prescription using the English ability evaluation/diagnosis /prescription system according to the present invention will be explained in detail .

FIG . 20 is a flow chart showing a process for evaluating English ability and offering an individualized diagnosis and prescription according to the present invention . As shown in FIG. 20, the process comprises a first step (SlOl) for outputting evaluation data classified according to sentence complexity and difficulty level, a second step (S102 ) for analyzing the structural complexity and difficulty level of sentences in the answer data inputted by the user in response to the evaluation data, a third step (S103) for calculating a correct answering rate in the answer data to evaluate the user' s English ability level and a fourth step (S104) for providing diagnosis and prescription concerning the user' s English ability based on the results of evaluation . The evaluation data consists of example sentences and questions which are outputted in text or speech .

The first step (SlOl) may output the evaluation data sequentially by stages classified according to the structural complexity and difficulty levels of sentences or randomly regardless of the stages . Different evaluation data can be used in different types of English ability evaluation test, for example, an English sentence construction test, an English sentence translation test and an English dialogue test . The first step (SlOl) further includes a sub-step of allowing the user to select a desired one of SCT (English

Sentence Construction Test) , STT (English Sentence Translation

Test) and DCT (English Dialogue Test) before outputting the evaluation data (i . e . , example sentences and questions) .

To this end, the English ability evaluation/diagnosis /prescription system provides the user with data necessary to select a desired test type through the text output means 260 or the speech output means 270. When the user selects a test type using the text input means 290 or the speech input means 280, evaluation data, such as example sentences and questions, corresponding to the selected test type are read out from the data storage unit 400 and outputted.

The example sentences and questions are outputted in text or speech according to the selected test type . For example, text example sentences and questions are outputted through the text output means 260. Preferably, such example sentences and questions should be displayed in a flash form that disappears after a few seconds . When the evaluation data (example sentences and questions) is outputted through the text output means 260 or the speech output means 270 according to the test type selected by the user, the user can input answer data in text through the text input means 290 or in speech through the speech input means 280. Whether the answer data should be inputted in text or in speech is determined according to the selected test type .

The answer data inputted by the user is transferred to the sentence ^• development stage classifier system 100 which will analyze the structural complexity and difficulty level of sentences included in the answer data. Based on the results of analysis, the sentence development stage classifier system 100 will determine the development stage of the answer data.

FIG. 21 is a flow chart showing a process for classifying sentence development stages according to the present invention. The process comprises the steps of converting a sentence forming the answer data which has been inputted in text or speech into computer-recognizable data (S201) , determining whether the sentence of the answer data is in a simple, semi-complex or complex pattern (S202) , classifying the sentence into a simple sentence development stage using a simple sentence complexity function if the sentence of the answer data is in a simple pattern (S203) , classifying the sentence into a semi-complex sentence development stage using a semi-complex sentence complexity function if the sentence of the answer data is in a semi-complex pattern (S204) , and classifying the sentence into a complex sentence development stage if the sentence of the answer data is in a complex pattern composed of at least two sentences, i . e . , a main clause and an embedded clause, by formulating a development stage of the main clause using the simple sentence complexity function or the semi-complex sentence complexity function and a development stage of the embedded clause using an embedded clause complexity function and combining the development stages of the main clause and the embedded clause together (S205) .

For example, in the following sentences, (a) Has he _ studied English?

(b) What are you _ studying?

(c) What does your father have _ you study _?

(d) Do you know _ what _ he has studied _ ? the bolded are constituents moved from the underlined positions, which means that there is a change of word order. As is clear from the above examples, interrogative sentences can be created by a grammatical change such as movement of constituents and a sentence pattern change such as word order change due to the movement of constituents .

A point of the present invention is to establish mechanisms of word order and movement of constituents that cause major grammatical and pattern changes and to quantify the relative complexity of English sentences in order to classify the development stages of the sentences . The three English construction mechanisms ("ECM") used in the present invention are [ Y SCM/ α CMM/J3 BOM] .

The three mechanisms, basic order mechanism ("BOM") , constituent movement mechanism ("CMM") and sentence combination mechanism ("SCM") , of the 3ECM will be explained in detail .

As a word order change mechanism in English sentences, BOM is defined by variable Jβ . If a sentence has no change in the basic word order [SVO] or [SVOC] of the simple or semi-complex pattern, it will be assigned a + value and defined as [+BOM] . On the other hand, a sentence having a change in the basic word order is assigned a - value and thus defined as [-B0M] .

CMM is a mechanism for movement (addition or insertion) of constituents necessary for a grammatical or pattern change of an English sentence. CMM is defined by variable α . A sentence with only one constituent moved (single movement) is assigned a + value and thus defined as [+CMM] . However, if two constituents are moved in a sentence (double movement) , in other words, if movement of one constituent causes another constituent to move, the sentence will be assigned a - value and thus defined [-CMM] .

Lastly, SCM is a sentence combination mechanism defined by variable y . In a complex sentence combining two or more sentences, i . e . , a main clause and an embedded clause, SCM is applied to the embedded clause. When two sentences are combined to construct a complex sentence, a constituent included in the embedded clause may be moved within the embedded clause . In such a case, the complex sentence is assigned a + value and thus defined as [+SCM] . On the other hand, if a constituent included in the embedded clause is moved to the main clause, a - value will be assigned to define the complex sentence as [-SCM] .

Hereinafter, each mechanism of 3ECM will be explained in more detail .

As explained above, the basic order mechanism [ Jβ BOM] assigns a + value or a - value to variable {3 according to whether the basic word order of S+V+O+C (subject+ verb + object + complement) is maintained or not . [Examples] (a) I love you. (b) You, I love _.

Of the above example sentences, (a) maintains the basic order of S+V+O and is therefore assigned 4- for variable β . However, (b) has a change of word order to O+S+V due to the movement of the object "you" to the beginning of the sentence , (b) is therefore assigned - for variable J3.

The - value is also assigned to Y/N-questions [V+S+O] formed by inverting the subject-verb order to the verb-subject order and [S+Aux+V+0] sentences having a change in the basic word order [S+V+O] due to the insertion of an auxiliary verb. As stated above, the basic order mechanism is correlated with the constituent movement mechanism.

The following are example sentences for explaining the constituent movement mechanism [ α CMM] . [Examples]

(a) [Will John _ marry Ann] next year?

(b) [Whom will John _ marry _] next year?

According to CMM, a sentence with only one constituent moved (single movement) is assigned a + value, whereas a sentence with two constituents moved (double movement) is assigned a - value.

(a) sentence is a Y/N question formed by moving only the auxiliary verb "will" to the beginning of the sentence (single movement) from the basic pattern [S+Will+V+O] . Therefore, (a) sentence is assigned + for variable α . The movement of only one constituent to achieve both a grammatical change and a pattern change is called a single movement .

In contrast, (b) sentence is a Wh-question formed by moving the interrogative "whom" corresponding to the object to the beginning of the sentence and thereby moving the auxiliary verb

"will" in front of the subject "John" (i . e . , inverting the order from subject-auxiliary verb to auxiliary verb-subject) . (b) question is thus assigned - for variable α .

The movement of two constituents to achieve both a grammatical change and a pattern change is called a double movement . Wh-questions, such as (b) , are typical examples of double movement .

In the above explanation of the constituent movement mechanism, the terms "movement" and "addition/insertion" are used. "Movement" refers to the relocation of a constituent originally included in a sentence . To be specific, when a constituent is moved from a starting point (an underlined "_" point in the above examples) to a destination point, it can be said that the constituent has been moved. "Addition" refers to the addition of a new constituent to a sentence . A nominative interrogative, such as who or what, is generally added to the beginning of a sentence to create a nominative interrogative sentence (for example, "Who kissed Ann?") .

Unlike (b) Wh-question, a nominative interrogative sentence formed by the addition of a nominative interrogative (Wh-word) to the beginning of the sentence does not entail inversion of the subject-auxiliary verb order. In other words, the addition of a Wh-word does not entail a change of word order. To the contrary, "insertion" of a grammatical constituent such as an auxiliary verb (can, may, have or the like) in the middle of a sentence (for example, "John will marry Ann") entails a change of word order.

The following are example sentences for explaining the sentence combination mechanism [ y SCM] . [Examples]

(a) [Do you know [who Mary likes _ in the class] ] ?

(b) [Who do you think [_ Mary likes _ in the class] ] ?

When two or more sentences are combined to form a complex sentence consisting of a main clause and an embedded clause, a constituent included in the embedded clause may move within the embedded clause or out from the embedded clause (i . e . , to the main clause) . SCM assigns a + value or a - value for variable Y according to whether a constituent in the embedded clause is moved within or out from the embedded clause . In Example (a) , the interrogative ^λλwho" corresponding to the object is moved to the beginning of the embedded clause (movement within the embedded clause) . Accordingly, (a) sentence is assigned + for variable y . In Example (b) , however, the interrogative "who" is moved to the beginning of the embedded clause and then to the beginning of the main sentence (movement out from the embedded clause) . Accordingly, (b) sentence is assigned - for variable

Y -

The three variables, α , J3 and Y , of 3ECM have been explained above. Although the variables of 3ECM can be used to quantify the structural complexity of English sentences, it is difficult to formulate the development stages of English sentences with only the variables . Hereinafter, it will be explained how the structural complexity of English sentences can be formulated to classify the development stages of the sentences .

As explained above, the three mechanisms of 3ECM are correlated with each another. Therefore, it is necessary to combine the three mechanisms to formulate the structural complexity of English sentences .

In view of the correlation of the three mechanisms and the structural characteristics of simple, semi-complex and complex sentences, the three mechanisms [ α CMM] , [ β BOM] and [ Y SCM] of

3ECM can be represented by the following combination functions to formulate the structural complexity of English sentences .

(a) Simple sentence/main clause complexity function: f{a , β ) = [ α CMM/β BOM] (b) Semi-complex sentence/embedded clause complexity function: f(α , β , γ ) = [Y SCM ( [ α CMM/β BOM] ) ] <Variables α , β and Y can be ^• , + or ->

Function (a) is a linear combination of [ α CMM] and [ β BOM] of 3ECM. Function (a) can be used to formulate the structural complexity of a simple sentence and a main clause of a complex sentence and thereby formulate the development stage of the simple sentence or the main clause .

Function (b) is an embed combination [Y SCM ( [ αCMM/β BOM] ) ] with function (a) (i . e. , the linear combination of [ α CMM] and [ β BOM] ) embedded in [ Y SCM] . Function (b) can be used to formulate the structural complexity of an embedded clause of a complex sentence or a semi-complex sentence and thereby formulate the development stage of the semi-complex sentence or the embedded clause . The logical basis of the embed combination (i . e. , function (b) with the linear combination of function (a) embedded in [ Y SCM] ) lies in that the complexity of a semi-complex or complex sentence is determined depending on the complexity of a simple sentence embedded in the semi-complex or complex sentence . The structural complexity of a complex sentence composed of a main clause and an embedded clause can be formulated by a combination of functions (a) and (b) , whereby the development stage of the complex sentence can be classified.

When + or - is assigned for variables α , J3 and Y of the above sentence complexity functions, the structural complexity of English sentences can be formulated according to the combination values of the variables in [ α CMM] / [ β BOM] and [Y SCM ( [ α CMM/β BOM] ) ] as shown in Table 2. Based on the formulated structural complexity, the development stages (or sequences) of English sentences can be classified. [Table 2]

As is clear from Table 2, the sentence combination mechanism [y SCM] is not applied to a simple sentence composed of a single sentence. Accordingly, simple sentences are classified into development stages IX, 2X, 3X and 4X through hierarchical combinations of variables α and β in the simple sentence complexity function f(a , β ) = [ α CMM/β BOM] .

A semi-complex sentence is structurally formed of a single sentence but is deemed to be composed of a main clause and an embedded clause in context . Accordingly, the sentence combination mechanism [Y SCM] is applied to a semi-complex sentence, without assigning a value for variable Y for the single sentence. Semi-complex sentences are classified into development stages 5Xa, 5Xb, 5Xc and 5Xd through hierarchical combinations of variables α and β in the semi-complex sentence complexity function f{a , β ) = [SCM ( [ α CMM/β BOM] ) ] which excludes variable Y from the embedded clause complexity function (b) , i . e . , f(a , β , γ ) = [Y SCM ( [ α CMM/β BOM] ) ] .

A complex sentence has a structure composed of a main clause and an embedded clause . To classify the development stage of a complex sentence, the embedded clause is first classified into development stage 1X2, 2X2, 3X2, 4X2 or 5X2 through hierarchical combinations of variables α , β and Y in the embedded clause complexity function f(a , β , y ) = [Y SCM ( [ α CMM/β BOM] ) ] and the main clause is classified into development stage IX, 2X, 3X or 4X through hierarchical combinations of variables α and β in the simple sentence complexity function _f(α , β ) = [ α CMM/β BOM] . The development stages of the embedded clause and the main clause are combined together to finally classify the development stage of the complex sentence .

Lastly, multiple-complex sentences in a structure composed of a main clause and two or more embedded clauses are classified using the same embedded clause complexity function and in the same manner as used to classify the complex sentences . All English sentences can be analyzed using the above ECM combination functions . The results of analysis will show that English sentences are classified into any of the development stages in Table 2. In the English sentence development stage classification as explained above, variables α and β in the ECM combinations, i . e. , [ αCMM/βBOM] , of simple sentences are applied in the ECM combinations of semi-complex or complex sentences, which means that the variables in the ECM combinations are in a hierarchical construction. In the development stages of English sentences, an upper stage sentence can include an ECM combination of a lower stage sentence according to the implicational scale . Conversely, a lower stage sentence cannot include an ECM combination of a higher stage sentence. For example, stage 2X can include an ECM combination of stage IX but cannot include a 3X combination or a critical structure thereof. The implicational scale implies that each development stage of English sentences is a gradual shedding of ECM combinations . Like a butterfly that grows from a caterpillar through shedding processes, each development stage of English sentences is reached after the shedding of the ECM combination of a previous stage. In other words, the ECM combination of each development stage should be shed along a predetermined development route in order to reach the next stage .

Hereinafter, it will be explained in detail how to formulate the structural complexity of English sentences to classify the development stages of the sentences . First, a process for classifying the development stages of simple sentences will be explained.

As explained above, simple sentences have a [SVO] pattern with a single tense verb. Since a simple sentence has only a single sentence, the sentence combination mechanism [γ SCM] cannot be applied. As shown in Table 2, four types of [ α CMM] / [ β BOM] combinations can be obtained by assigning + or - to variables α and _β in the simple sentence complexity function f{a , β ) = [ α CMM] / [ β BOM] .

When the structural complexity of simple sentences are quantified according to the values of the ECM combinations, the simple sentences can be classified into development stages IX, 2X, 3X and 4X based on the quantified structural complexity.

The first development stage IX of simple sentences is represented by [ ^• CMM] / [+B0M] of the ECM mechanism. Simple sentences in this stage have neither a movement of constituent

( [ ^• CMM] ) nor a change of word order ( [+B0M] ) as compared to the standard SVO pattern.

The second development stage 2X is represented by

[+CMM] / [+BOM] of the ECM mechanism. Simple sentences in this stage have no change in grammar or word order ( [+BOM] ) despite a constituent moved or added to the beginning of the SVO sentence pattern ( [+CMM] ) .

The third development stage 3X is represented by [+CMM] / [- BOM] of the ECM mechanism. Simple sentences in this stage have a change in word order ( [-B0M] ) due to a new constituent added to the sentence or a constituent moved to the beginning from the middle of the SVO sentence pattern ( [+CMM] ) . The fourth development stage 4X is represented by [-CMM] / [- BOM] of the ECM mechanism. An SVO simple sentence in this stage has a double movement (i . e. , a movement of a constituent to the beginning of the sentence and a consequent movement of another constituent within the sentence) ( [-CMM] ) , which results in a change in word order ( [-BOM] ) .

In summary, simple sentences in English are classified into development stages IX, 2X, 3X and 4X according to hierarchical combinations of variables α and β in [ α CMM/β BOM] . Accordingly, any simple sentence in SVO pattern can be classified into one of the development stages IX, 2X, 3X and 4X.

In the English sentence development stage classification as explained above, variables α and β in the ECM combinations

( [ α CMM/β BOM] combinations) of simple sentences are in a hierarchical construction. Accordingly, an upper stage sentence can include an ECM combination of a lower stage sentence according to the implicational scale . Conversely, a lower stage sentence cannot include an ECM combination of a higher stage sentence. For example, stage 2X can include an ECM combination of stage IX but cannot include a 3X combination or a sentence structure thereof.

Hereinafter, each development stage of simple sentences in English will be explained in detail . 1. Development Stage IX

Stage IX is a starting point of the development stages of English sentences . In Example 1,

[Example 1] (a) Go out !

(b) Be good. Kiss me.

(c) I go there .

(d)^' I am a boy. I kissed my mother at the park.

Sentences (a) through (d) have neither a movement of constituent nor a change of basic word order. For these sentences, variable α in the simple sentence complexity function f(α , J3 ) = [ α CMM] / [ β BOM] is assigned ^• (none) because of no movement of constituent, while variable J3 is assigned + because of no change of word order. Accordingly, [ ^• CMM] / [+BOM] of the ECM mechanism is applied to sentences (a) through (d) to define the sentences as stage IX. The critical structure governed by the mechanism [ ^• CMM] / [+BOM] is [ (S) V (P) (0) ] (Adv) as shown in

Example 1. Constituents in the parentheses ( ) are optional .

Conclusively, sentences classified into development stage IX under [ - CMM] / [+BOM] of the ECM mechanism have a basic SVO pattern with no change in word order.

2. Development Stage 2X

The following sentences of Example 2 have a movement or addition of constituent but no change in word order. [Example 2]

(a) At the park, [I kissed Mary] _ last night .

(b) *Do [you kissed Mary] last night? (c) *røhere [John Kissed Mary] _ last night? {Where did John kiss Mary last night?) (* is a sentence with a grammatical error. )

The bolded are constituents moved from the underlined to the beginning of the sentence or newly added to the beginning of the sentence . The movement or addition of a constituent in each of the above sentences does not entail a change of basic word order [ (S) V (O) ] . For the above sentences, variable α in the simple sentence complexity function r^~(α , β ) = [ α CMM] / [ Jβ BOM] is assigned + because a constituent is moved or added to the beginning of each sentence. Variable J3 is also assigned + because such movement or addition of constituent does not result in a change of word order.

Accordingly, [+CMM] / [+BOM] of the ECM mechanism is applied to define the above sentences as stage 2X. No change of word order means that constituents added or moved (i . e . , "At the part, " "Do" and "Where") do not grammatically affect the other constituents in the [ (S) V (O) ] sentences as is clear from Example

2. For example, if the addition of "Do" to the beginning of sentence (b) grammatically affects the other constituents in the sentence, "Do" must be changed to a past tense form "Did" and

"kissed" to a root form "kiss" at the same time . In sentence

(c) , "Where" is moved to the beginning from the middle of the sentence . Since the moved constituent does not grammatically affect the other constituents in the sentence, the subject and verb order in the sentence is not inverted. Accordingly, sentence (c) is a grammatically incorrect sentence .

Unlike typical analyses of English sentences, the English development stage classification according to the present invention does not seriously consider grammatical errors in English sentences . Typically, grammatically erroneous sentences are even excluded from an English language analysis . Also, People who frequently makes grammatical errors are deemed to have lower level language skills .

In the English development stage classification according to the present invention, grammatical errors are considered to naturally occur at each English development stage . Moreover, grammatical errors resulting from the application of the ECM mechanism at each development stage (for example, sentences (b) and (c) in Example 2) are considered as being absolutely necessary to reach the next development stage. Actually, people who make grammatical errors such as sentences (b) and (c) can master each English development stage in a shorter time than those who do not .

The English development stage classification according to the present invention can predict which kinds of grammatical errors are likely to be made at a specific development stage and accurately analyze whether a grammatical error in a sentence is a systematic error resulting from the application of the ECM mechanism or an unsystematic error resulting from a mere mistake to determine the pertinent English development stage .

Conclusively, sentences classified into development stage 2X under [+CMM] / [+BOM] of the ECM mechanism have a fronting structure that simply adds or moves an adverb phrase (Adv) , auxiliary verb (D) or Wh-word (Who, What, When or the like) to the beginning of a critical sentence structure of stage IX. Since a 2X sentence is not governed by [-BOM] mechanism, addition or movement of a constituent to the beginning of the sentence does not grammatically affect the other constituents in the sentence .

3. Development Stage 3X

The following sentences of Examples 3, 4 and 5 have a movement of constituent and a change in basic word order.

[Example 3] (a) I can play tennis .

(b) I want to meet her.

(c) I enjoy watching action movies . [Example 4]

(a) I am watching TV now. ( [S be+ (not) +V-ing (0) ] ) (b) Mary was killed. ( [S be+ (not) +V-en (0) ] )

(c) John has not written a book. ( [S have+ (not) +V-en (0) ] )

(d) John has been reading the novel . ( [S have/be+ (not) + V-en/-ing (0) ] )

[Example 5] (a) Can you _ play tennis this weekend? (M [S_V (O) ] )

(b) Did you kiss_ Mary last night? (Do [SV_(0) ] )

(c) Would you _ mind smoking here? (M [S_V+V-ing (0) ] )

(d) Have you _ written the essay? (A [S_V-en (O) ] )

In Example 3, a modal (can, may, must or the like) or to+V or V-ing is inserted into an [SVO] sentence, thereby changing the basic word order [SVO] to [S M+V+toV/+V-ing 0] .

In Example 4, a grammatical constituent (auxiliary verb) for expressing a verb tense such as present perfect (have+pp) , present continuous (be+ing) or passive (be+pp) is inserted into an [SVO] sentence (in front of or after a main verb) , thereby changing the basic word order [SVO] to [S Aux+V+I 0] .

In Example 5, unlike in Examples 3 and 4 , a modal or an auxiliary verb (such as "can, " "did, " "would" or "have") included in the middle (underlined position) of an [SVO] sentence is moved to the beginning of the sentence to form a Y/N-question, thereby causing a change in the basic word order [SVO] .

In view of the insertion of constituent and the change of word order, the simple sentence complexity function f{a , β ) =

[ α CMM] / [ J3 BOM] can be determined. For each sentence in Examples

3 to 5, variable α in the simple sentence complexity function is assigned + because a grammatical constituent (modal/auxiliary verb) is inserted into each [SVO] sentence or moved to the beginning from the middle of the sentence . Variable J3 is assigned - because such insertion or movement of constituent entails a grammatical change or a word order change .

Accordingly, [+CMM] / [-B0M] of the ECM mechanism is applied to define the above sentences as stage 3X. To further explain a grammatical change and a word order change in an [ ( s) V (O) ] sentence, sentence (b) in Example 5 will be compared with sentence (b) in Example 2 (development stage 2X) .

(b) of Example 2 :

*Do [you kissed Mary] last night? [+CMM] / [+BOM] (b) of Example 5 :

Did you kiss_ Mary last night? [+CMM] / [-BOM] As explained above, the constituent "Do" added to the beginning of sentence (b) of Example 2 does not grammatically affect the verb "kissed" in the [ (S) V (O) ] sentence . Therefore, the addition of constituent does not entail a grammatical change such as "Do" to a past tense form "Did" and "kissed" to a root form "kiss . " In contrast, the auxiliary verb "Did" added (moved) to the beginning of sentence (b) of Example 5 entails a grammatical change that deletes "-ed" from "kissed" to change the verb to a root form "kiss . " The two sentences having a constituent added or moved are both assigned [+CMM] . However, one (sentence (b) of Example 2) having no grammatical change or word order change is assigned [+BOM] , whereas the other (sentence (b) of Example 5) having a grammatical change and a word order change is assigned [-BOM] . The grammatical error ("kissed") in sentence (b) of Example 2 is not a systematic error resulting from the application of the [-BOM] but an unsystematic error resulting from a mere imitation to form an English sentence .

[+BOM] and [-B0M] are determined depending on whether the basic word order is changed or not . In actual English language proficiency levels, + and - values can make a great difference. People having no ability to make an appropriate change of word order in an SVO pattern are deemed to lack understanding of grammatical relationship between the sentence constituents . On the other hand, those who can appropriately change the basic word order are deemed to fully understand the grammatical relationship between the constituents and have an ability to make appropriate grammatical changes and basic order changes in English sentences . The critical structures of development stage 3X governed by [+CMM] / [-B0M] are as shown in Table 3. [Table 3]

Modal Verbs : can, may, must, etc .

Auxiliary Verbs: be, have, do and other auxiliary verbs used for verb tenses such as present perfect/present continuous /passive .

Inflections : verb endings such as "-ed, " ^λl-es, " "-en" and "-ing"

Structure A has a grammatical element such as M, I , A, to+V or V-ing is inserted in the SVO pattern, particularly before or after V (verb) , and includes four sub-structures a, b, c and d as shown in Table 3 . Of the four sub-structures , "d" is the most complex structure because it has two grammatical elements inserted in the SVO pattern . Structures including "be going to do, " "have to do" or "be able to do" are in the same complexity- level . Structure B above moves an inserted modal or auxiliary verb to the beginning of the sentence to form a Y/N-question .

4 . Development: Stage 4X

It will be discussed how [ α CMM] / [ J3 BOM] of ECM works in the sentences of Examples 6 and 7 .

[Example 6] Whom will you invite to the party?

( a) [You will invite whom] to the party?

(b) [Whom You will invite ] to the party? (c) [Whom will You _ invite _] to the party? [Example 7]

Never would I talk to his sister about this .

(d) I would never talk to his sister about this . The Wh-question of Example 6 is completed through the structuring of sentences (a) and (b) . "Whom" corresponding to the object in SVO structure (a) is moved to the beginning of the sentence to form structure (b) . The movement of "Whom" causes another constituent "will" in SVO to move before the subject "You" (in other words, inversion of subject-verb to verb- subject) , thereby changing the word order from [Wh S M V] to [Wh M S V] . Example 6 shows a double movement which means that a movement of a grammatical constituent (such as a Wh-word) to the beginning of a sentence entails a consequent movement of another constituent in the sentence .

Example 7 is a so-called Neg-inversion structure . A negative "never" in sentence (d) is moved to the beginning of the sentence, thereby resulting in a subsequent movement, i. e . , inversion of the subject-auxiliary verb order "I would" to the auxiliary verb-subject order "would I . " Due to the double movement, the basic SVO word order is changed.

In view of the movement of constituent and the change of word order, the simple sentence complexity function Jf(α , β ) = [ α CMM] / [ β BOM] can be determined. For the complete sentences in Examples 6 and 7, variable α in the simple sentence complexity function is assigned - because of the double movement of constituents . Variable Jβ is also assigned - because the double movement of constituents entails a grammatical change and a word order change . Accordingly, [-CMM] / [-B0M] of the ECM mechanism is applied to define the above sentences as stage 4X. In other words, Wh-question or Neg-question structures having a double movement of constituents are governed by [-CMM] / [-BOM] . Conclusively, sentences classified into stage 4X under [- CMM] / [-BOM] of ECM have a double movement (i . e . , a movement of a constituent to the beginning of the sentence and a consequent movement of another constituent in the sentence) which entails a word order change and a grammatical change .

As explained above, the 3ECM mechanisms and combination functions can be used to evaluate the ability to make a grammatical change and a pattern change through movement of constituents and change of basic word order . Particularly, the development stages of simple sentences in English can be classified into IX to 4X through hierarchical combinations of variables α and β in the simple sentence complexity function jT(α , β ) = [ α CMM/jβ BOM] . The ECM mechanism can be better understood through Example 8. [Example 8]

(a) Who [kissed Mary] last night? (a' ) Who [_kissed Mary] last night?

(b) Wham did [John kiss ] last night?

Sentences (a) and (b) are both Wh-questions formed by the movement (addition) of a Wh-word. However, the two sentences are different in the type of movement and the basic order change which are related to the structural complexity and difficulty level of English sentences . The structural complexity and difficulty level can be determined by the ECM mechanism. According to the movement of constituent and word order change captured by the ECM mechanism, variables α and Jβ are assigned a + value or a - value, thereby quantifying the relative complexity of a specific English sentence . Then the variable values + or - are applied to the simple sentence complexity function f(α , Jβ ) = [ α CMM/β BOM] to finally classify the development stage of the simple sentence .

The development stages of sentences (a) and (b) in Example 8 can be classified in the manner explained above .

In sentence (a) (Who [kissed Mary] last night?) , "who" is deemed to have been added to the basic [ (S) V (O) ] sentence, i . e . , to the beginning of the sentence [kissed Mary] , or to have been moved to the beginning of the sentence [_ kissed Mary] from the underlined position. Despite the addition or movement of "who, " the basic word order is maintained in SVO pattern.

According to the ECM, addition of a constituent "who" is regarded a being the same as movement of the constituent . Sentence (a) of Example 8 clearly has a movement of a constituent but no change in basic word order. In other words, the addition or movement of a constituent does not entail a grammatical change (for example, a change in the verb ending) in the basic SVO pattern.

To quantify the structural complexity of sentence (a) , variable α in [α CMM] ("Constituent Movement Mechanism") is assigned + because there is a single movement of constituent . Also, variable Jβ in [ JβBOM] is assigned + because there is no word order change. When the variable values + or - are applied to the simple sentence complexity function f(a , β ) = [ α CMM/β BOM] , [+CMM] / [+BOM] is obtained for sentence (a) .

Conclusively, sentence (a) of Example 8 is classified into development stage 2X. In sentence (b) of Example 8 (Whom did [John kiss_ _] last night?) , "whom" corresponding to the object of "kiss" in the basic [ (S) V (O) ] sentence is first moved to the beginning of the sentence to form a new sentence "Wham [John kissed _] last night?" With the movement of "whom, " the past-tense ending "-ed" of the verb "kissed" is replaced with an auxiliary verb "did" which is then moved before the subject "John. " According to the inversion of the subject-verb order, sentence (b) "Whom did [John kiss ] last night?" is completed. In other words, sentence (b) has a double movement which means that the movement of "whom" to the beginning of the sentence causes a subsequent movement of the auxiliary verb "did" .

With the movement of "whom" which corresponds to the object O in the basic [S V+ed O] sentence, the word order is changed to [Whom S V+ed ] . The word order is changed again to [Whom d±d(_ed) S V_ _] by the replacement of the past-tense ending "- ed" with the auxiliary verb "did. "

For sentence (b) , variable α in [α CMM] ("Constituent Movement Mechanism") is assigned - because there is a double movement of constituents . Also, variable β in [ β BOM] is assigned - because there is a word order change . When the variable values are applied to the simple sentence complexity function f(a , J3 ) = [ α CMM/β BOM] , [-CMM] / [-BOM] is obtained for sentence (b) . Therefore, sentence (b) of Example 8 is classified into development stage 4X.

According to traditional English grammar, sentences (a) and

(b) in Example 8 are both Wh-questions . However, according to the ECM that analyzes the movement of constituent and the basic order change, sentence (b) is two times more complex than sentence (a) . Consequently, sentence (a) is classified into 2X and sentence (b) into 4X. Actually, English learners are likely to make more grammatical errors in sentences having the structure of (b) , rather than (a) . Many of them can only compose an intermediate sentence "Wham [John kissed _] last night?" by moving the "Wh-word" to the beginning of the sentence, without completing the sentence to "Whom did [John kiss ] last night?" through a subsequent movement . Those who lack understanding of the double movement mechanism, i . e . , [-CMM] , often use such grammatically incorrect sentences .

Unlike traditional English grammar that merely explains the grammar rules in English, the English sentence development stage classification according to the present invention focuses on a "change" in "grammar or pattern" of sentences to accurately evaluate a learner' s English language proficiency and predict possible grammatical errors . Therefore, the English sentence development stage classification according to the present invention can provide each learner with accurate diagnosis as to what (development stage) to study to improve English language skills . Also, the classification can be of great help in overall English education (instruction, . learning and evaluation) and improve the language learning efficiency.

The method of classifying the development stages of simple sentences as explained above can be used in the classification of semi-complex, complex or multiple-complex sentences .

A semi-complex sentence is structurally formed of a single sentence but is deemed to be composed of a main clause and an embedded clause in context . Accordingly, the sentence combination mechanism [y SCM] is applied to a semi-complex sentence, without assigning a value for variable Y for the single sentence . Semi-complex sentences are classified into development stages 5Xa, 5Xb, 5Xc and 5Xd through hierarchical combinations of variables α and β in the semi-complex sentence complexity function f(α , β ) = [SCM ( [ α CMM/β BOM] ) ] which excludes variable y from the embedded clause complexity function (b) , i . e. , f(α , β , Y ) = [Y SCM ( [ α CMM/β BOM] ) ] . The structural complexity of semi-complex sentences is quantified based on the ECM combination values, thereby classifying the semi-complex sentences into the four development stages 5Xa, 5Xb, 5Xc and 5Xd.

The same scale values of [ α CMM/β BOM] of ECM at stages IX,

2X, 3X and 4X are applied to [ αCMM/β BOM] at stages 5Xa, 5Xb,

5Xc and 5Xd (see variables in the simple sentence complexity function) . In other words, the mechanism for classifying simple sentences into development stages IX, 2X, 3X and 4X is used in the mechanism for classifying semi-complex sentences into development stages 5Xa, 5Xb, 5Xc and 5Xd. It is possible to explain how semi-complex sentences are classified into stages 5Xa, 5Xb, 5Xc and 5Xd based on the mechanism of classifying simple sentences into stages IX, 2X, 3X and 4X. Semi-complex sentences and simple sentences constitute an implicational scale . 1. Development Stage 5X

The following sentences will be classified using the semi- complex sentence complexity function _f(α , β ) = [SCM ( [ α CMM /β BOM] ) ] . (a) [My father wants me to marry the millionaire] .

(b) Who [wants you to marry the millionaire] ?

(c) [Does your father want_ you to marry the millionaire] ?

(d) [Who does your father want_ you to marry _] ? Sentence (a) has neither a change in the semi-complex sentence pattern [SVOC] nor a change in the basic word order. Therefore, variables Q and β in the semi-complex sentence complexity function f(a , β ) = [SCM ( [ α CMM/β BOM] ) ] are assigned " ^• " (none) and "+, " respectively. As a result, [SCM ( [ ^• CMM] / [- BOM] ) ] of the ECM mechanism is applied to define sentence (a) as stage 5Xa . Critical structures of this stage include [SVOC] like sentence (a) and [SVOO] which is a so-called double object structure .

Sentence (b) is a Wh-fronting structure having a nominative interrogative "who" added to the beginning of the semi-complex sentence pattern [SVOC] . The addition of "who, " however, does not causes a change in the basic order of [SVOC] . In other words, unlike a Wh-question at stage 4X, sentence (b) has no subject-verb inversion. Therefore, variables α and β in the semi-complex sentence complexity function f{& , β ) = [SCM ( [ α CMM/β BOM] ) ] are both assigned "+. " As a result, [-SCM ( [+CMM] / [+BOM] ) ] of the ECM mechanism is applied to define sentence (b) as stage 5Xb. A nominative interrogative sentence structure (Wh-fronting: Wh [V 0 OC] ) is the critical structure of this stage .

Sentence (c) is a Y/N-question structure formed by changing the inflection "-es" of the verb in [SVOC] to an interrogative auxiliary verb "does" and moving the auxiliary verb "does" to the beginning of the sentence . Sentence (cc) has a grammatical change such as a change of the verb ending (i . e . , change of the verb to a root form) and a change of word order. Therefore, variables α and J3 in the semi-complex sentence complexity function f(α , β ) = [SCM ( [ α CMM/β BOM] ) ] are assigned "+" and "- , " respectively. As a result, [-SCM ( [+CMM] / [-B0M] ) ] of the ECM mechanism is applied to define sentence (c) as stage 5Xc. A Y/N question (M/A S _V0C) is the critical structure of this stage .

In sentence (d) , "whom" corresponding to the object of the verb "marry" is moved to the beginning of the sentence, which causes the auxiliary verb "does" corresponding to the inflection "-s" of the verb "wants" in [SVOC] to move before the subject (subject-auxiliary verb inversion) . Due to the double movement of constituents, the basic word order is changed. Therefore, variables α and J3 in the semi-complex sentence complexity function _f(α , β ) = [SCM ( [ α CMM/β BOM] ) ] are both assigned "- . " As a result, [-SCM ( [-CMM] / [-BOM] ) ] of the ECM mechanism is applied to define sentence (d) as stage 5Xd. A Wh-question like sentence (d) [Wh M/A S _V (O) (C) ] is the critical structure of this stage . It has been explained above how to classify the development stages of English simple and semi-complex sentences according to the present invention. Hereinafter, a method of classifying the development stages of complex sentences will be explained in detail .

A complex sentence has a structure composed of at least two simple or semi-complex sentences, i . e . , a main clause and an embedded clause. First, the development stage of a main clause can be classified using the simple sentence complexity function Jr (Ct , β ) = [ α CMM/β BOM] (when the main clause is a simple sentence) or the semi-complex sentence complexity function f(α , β ) = [SCM ( [ α CMM/β BOM] ) ] (when the main clause is a semi-complex sentence) . Embedded clauses can be classified into five development stages IX², 2X², 3X², 4X² and 5X² by obtaining five types of [SCM ( [CMM] / [BOM] ) ] combination from hierarchical combinations of variables α , β and y in the embedded clause complexity function JT(Ci , β , Y ) = [Y SCM ( [ α CMM/βBOM] ) ] and quantifying the structural complexity of embedded clauses in English based on the five ECM combination values .

The same scale values of [ α CMM/β BOM] of ECM at stages IX, 2X, 3X and 4X are applied to [ α CMM/β BOM] at stages IX², 2X², 3X², 4X² and 5X². In other words, the mechanism for classifying simple sentences into the development stages IX, 2X, 3X and 4X is used in the mechanism for classifying embedded clauses into the development stages IX², 2X², 3X², 4X² and 5X².

It is possible to explain how semi-complex sentences are classified into stages 5Xa, 5Xb, 5Xc and 5Xd, or how embedded clauses are classified into stages IX², 2X², 3X², 4X² and 5X² based on the mechanism of classifying simple sentences into stages IX, 2X, 3X and 4X. Embedded clauses, semi-complex sentences and simple sentences constitute an implicational scale . Unlike semi-complex or simple sentences, embedded clauses has one more development stage 5X². ^' A complex sentence with a constituent moved out from an embedded clause during the combination of two simple sentences, i . e . , a main clause and the embedded clause, is classified into the fifth development stage 5X².

When the structural complexity of an embedded clause is formulated in such a way, only the movement ( [ ±CMM] ) of a critical factor associated directly with the embedded clause (a connective, conjunction, interrogative, relatival or the like) and the change in word order ( [ +BOM] ) will be considered to apply the ECM mechanism of the embedded clause and obtain variables α , β and Y in [y SCM ( [ α CMM/β BOM] ) ] .

Therefore, the structure of the embedded clause itself, for example, whether the embedded clause is an SVO pattern or an SVOOC pattern, will not be further considered to formulate the complexity of the embedded clause. According to the implicational scale, a learner who knows the structure of embedded clauses is deemed to have already learned the structures of simple and semi-complex sentences .

To determine the development stage of a complex sentence which is composed of a main clause and an embedded clause, the development stage of the embedded clause is first formulated using the simple sentence complexity function and the semi- complex sentence complexity function and then the development stages of the main clause and the embedded clause are combined together. Therefore, the development stage of a complex sentence can be represented by "KX² + kx. " For example, "3X² + 2x" stage of a complex sentence implies that the main clause and embedded clause of the complex sentence are of development stages 2x and 3X², respectively.

1. Development Stage IX²

Any movement of constituent or change of word order will be considered in the following sentence of Example 9. [Example 9]

(a) I think (that) you can help him to do his homework. (b) [SCM] : [I think + [ { that) you can help him to do his homework. ] ]

As shown in (b) [SCM] (Sentence Combination Mechanism) ,

Example 9 is a complex sentence having a main clause and an embedded clause which are combined together using a conjunction "that . " The combination of two sentences (main and embedded clauses ) does not entail a change of word order of the embedded clause "you can help him to do his homework. " Therefore, in the embedded clause complexity function Jf(Ci , β , Y ) =

[Y SCM ( [ α CMM/J3 BOM] ) ] , variables α and J3 are assigned " ^• " (none) and "+, " respectively. Since there is no movement of constituent within the embedded clause, variable Y is assigned

" ^• " (none) . As a result, [-SCM ( [ ^• CMM] / [+BOM] ) ] of the ECM mechanism is applied to define the embedded clause of Example 9 as stage IX². Since the main clause "I think" is maintained in the basic order [SVO] , variables α and β in the simple sentence complexity function f(α , β ) = [ α CMM/β BOM] are assigned " ^• " (none) and "+, " respectively. Thus, [ ^• CMM/+BOM] of the ECM mechanism is applied to define the main clause of Example 9 as stage IX.

Consequently, the complex sentence of Example 9 is classified into development stage IX² + Ix which means that the embedded clause and main clause of the sentence are of development stages IX² and IX, respectively.

2. Development Stage 2X²

Any movement of constituent or change of word order will be considered in the following sentence of Example 10. [Example 10]

(a) Do you know who taught English to this class last year?

(b) [SCM] : [Do you know + [who taught English to this class last year] ] ?

As shown in (b) [SCM] (Sentence Combination Mechanism) , Example 10 is a complex sentence having a main clause and an embedded clause which are combined together using an interrogative "who. " In the embedded clause, "who" (fronting; nominative interrogative) is added (moved) to the beginning of the clause, without changing the basic word order. Therefore, variables α and J3 in the embedded clause complexity function f(α , β , Y ) = [Y SCM ( [ αCMM/β BOM] ) ] are both assigned "+. ". Since the constituent "who" is added (moved) within the embedded clause, variable Y is also assigned "+. " Therefore, [+SCM ( [+CMM] / [+B0M] ) ] of the ECM mechanism is applied to define the embedded clause of Example 10 as stage 2X².

The main clause "Do you know" is formed by moving the auxiliary verb "do" which is deemed to be a constituent included in [SVO] to the beginning of the sentence . The movement of "do" causes a change in the basic word order [SVO] . Therefore, variables α and β in the simple sentence complexity function f{a , β ) = [ α CMM] / [ J3 BOM] are assigned "+" and "-, " respectively. Under [+CMM/-B0M] of the ECM mechanism, the main clause of Example 10 is classified as stage 3X. Consequently, the complex sentence of Example 10 is classified into development stage 2X² + 3x. [Example 11]

(a) Have you ever met the lady who was talking with John in the park yesterday?

(b) Do you know the time when he will come?

(c) While the show was holding, the accident happened. Unlike the indirect question of Example 10 using a nominative interrogative, sentences (a) , (b) and (c) of Example 11 use a nominative relative, relative adverb and adverb clause (time) , respectively. Sentences (a) to (c) have a Wh-word fronting to the beginning of their embedded clauses . The three sentences (a) to (c) are similar to the indirect question of Example 10 in that the Wh-word fronting causes neither a grammatical change nor a word order change. Therefore, the same mechanism [+SCM ( [+CMM] / [+BOM] ) ] as used in Example 10 is applied to define the embedded clauses of Example 11 as stage 2X².

3. Development Stage 3X²

Any movement of constituent or change of word order will be considered in the following sentence of Example 12. [Example 12]

(a) Have you ever met the man whom my father expects me to marry __?

(b) [SCM] : [Have you ever met the man + [whom my father expects me to marry _?] ]

As shown in (b) [SCM] , Example 12 is a complex sentence using an objective relative pronoun "whom" to combine two sentences, i . e . , a main clause and an adjective clause . In the adjective (embedded) clause of Example 12 ("whom my father expects me to marry _") , "whom" corresponding to the object of the verb "marry" is moved to the beginning of the clause, thereby causing a change of word order in the clause. Therefore, variables α and β in the embedded clause complexity function f(α , β , Y ) = [y SCM ( [ α CMM/β BOM] ) ] are assigned "+" and "-, " respectively. Since the movement of constituent "whom" is made within the embedded clause, variable Y is also assigned "+. " As a result, [+SCM ( [+CMM] / [-BOM] ) ] of the ECM mechanism is applied to define the embedded clause of Example 12 as stage 3X².

The main clause "Have you ever met the man ... ?" is formed by moving the auxiliary verb "have" which is deemed to be a constituent included in [SVO] to the beginning of the sentence. The movement of "have" causes a change in the basic word order [SVO] . Therefore, variables α and β in the simple sentence complexity function _?(α , β ) = [ α CMM] / [ β BOM] are assigned "+" and "-, " respectively. Under [+CMM/-B0M] of the ECM mechanism, the main clause of Example 12 is classified as stage 3X. Consequently, the complex sentence of Example 12 is classified into development stage 3X² + 3x. 4. Development Stage 4X²

Any movement of constituent or change of word order will be considered in the following sentence of Example 13. [Example 13]

(a) Do you ask Mary whom _ she will invite _ to the party?

(b) [SCM] : [Do you ask Mary + [whom will she _ invite _ to the party] ] ?

As shown in (b) [SCM] , Example 13 is a so-called indirect question (objective) in a complex sentence structure that combines two sentences (a main clause "Do you ask Mary ...?" and an embedded clause "whom will she _ invite _ to the party?") using an interrogative "whom. "

In Example 13, the direct question "whom will she _ invite _ to the party?" was converted to the embedded clause (indirect interrogative clause) . In the embedded clause, "whom" corresponding to the object of the verb "invite" is moved to the beginning of the clause, which causes the auxiliary verb "will" to move before the verb "invite" in the clause. Due to this double movement of constituents, the basic word order is changed.

Therefore, variables α and β in the embedded clause complexity function f(a , J3 , y ) = [y SCM ( [ α CMM/β BOM] ) ] are both assigned "- //

Since the movement of constituents is made within the embedded clause, variable y is assigned "+ . " As a result, [+SCM ( [-CMM] / [-BOM] ) ] of the ECM mechanism is applied to define the embedded clause of Example 13 as stage 4X².

The main clause "Do you ask Mary ...?" has the semi-complex sentence pattern [SVOC] . The auxiliary verb "do" which is deemed to be a constituent included in the [SVOC] pattern is moved to the beginning of the sentence, thereby causing a change in the basic word order. Therefore, variables α and jβ in the main clause complexity function (in this case, semi-complex sentence complexity function) f(a , β ) = [SCM ( [ α CMM/β BOM] ) ] are assigned "+" and "-, " respectively. Under [SCM ( [+CMM/-B0M] ) ] of the ECM mechanism, the main clause of Example 13 is classified as stage 5Xc. Consequently, the complex sentence of Example 13 is classified into development stage 4X² + 5Xc which means that the embedded clause and the main clause are of development stages 4X² and 5Xc, respectively.

5. Development Stage 5X² Any movement of constituent or change of word order will be considered in the following sentence of Example 14. [Example 14] (a) [ [Wham [do you think _] [_ my father _ wants me to marry _] ] ? (b) [SCM] : [do you think] + {Wham my father _ wants me to marry _} ] ?

(i) <τ Wham does my father want_ me to marry _? (ii) <- Wham _ my father wants me to marry _? (iii) ^- [ [Wham [do you think _] [_ my father _ wants me to marry _] ] ?

Sentence (a) of Example 14 is a so-called long-distance wh- question having a Wh-word moved to the main clause from the embedded clause. As shown in (b) [SCM] , two sentences (a main clause and an embedded clause) are combined together by means of an interrogative "whom" to form a complex sentence. Specifically, the direct question (i) "whom does my father want me to marry _?" was converted to the embedded clause (ii) (indirect interrogative clause) by moving "whom" corresponding to the object of the verb "marry" to the beginning of the clause and changing the auxiliary verb "does" to an inflection "-s" to be moved after the verb "wants" (double movement of constituents) . When the main clause uses an opinion verb such as "think, " "believe, " "suppose" or "imagine, " the interrogative "whom" at the beginning of the embedded clause (ii) is moved again to the beginning of the main clause to complete the complex sentence structure (iii) . In view of the above double movement of constituents and change of word order in the embedded clause, variables α and _β in the embedded clause complexity function f{a , J? , Y ) = [Y SCM ( [ α CMM/β BOM] ) ] are both assigned "- . "

Since the interrogative "whom" is moved to the beginning of the main clause from the embedded clause, variable Y is assigned

"- . " As a result, [-SCM ( [-CMM] / [-BOM] ) ] of the ECM mechanism is applied to define the embedded clause of Example 14 as stage 5X².

The main clause "do you think ...?" has the semi-complex sentence pattern [SVOC] . The auxiliary verb "do" which is deemed to be a constituent included in the [SVO] pattern is moved to the beginning of the sentence, thereby causing a change in the basic word order. Therefore, variables α and J3 in the main clause complexity function (in this case, simple sentence complexity function) f(α , Jβ ) = [ α CMM] / [ Jβ BOM] are assigned ^ΛΛ+" and "-, " respectively. Under [+CMM/-B0M] of the ECM mechanism, the main clause of Example 14 is classified as stage 3X. Consequently, the complex sentence of Example 14 is classified into development stage 5X² + 3x.

6. Development Stage X³

X³ is the development stage corresponding to a multiple- complex sentence composed of two or more embedded clauses . The embedded clause complexity function is applied to classify the development stages of multiple-complex sentences in the same manner as used in the complex sentence classification. Any movement of constituent or change of word order will be considered in the following sentence of Example 15. [Example 15] [Who [do you guess [_ Mary thinks [_ John loves _ at the college] ] ] ?

Example 15 is a multiple-complex sentence with two embedded clause, i .e . , a first embedded clause and a second embedded clause embedded in the first embedded clause. Since the same mechanism as used to classify the development stages of complex sentences can be used to classify the development stage of the above multiple-complex sentence, only the movement of the interrogative "who" will be explained. The interrogative "who" corresponding to the object of the verb "loves" is first moved to the beginning of the first embedded clause, thereby changing the first embedded clause to [who Mary thinks ...] . The interrogative "who" is moved again to the beginning of the main clause, thereby changing the main clause to [Who do you guess ...] . In view of such movements of constituent "who" and changes of word order, the multiple-complex sentence of Example 15 is classified into development stage 5X³ + 5X² +3x which means that the development stages of the second embedded clause, first embedded clause and main clause are 5X³, 5X² and 3x, respectively. A multiple- complex sentence may have two embedded clauses like Example 15, or three or more embedded clauses . Any multiple-complex sentence with n embedded clauses can be classified using the embedded clause complexity function. This is one of the advantages of ECM. As explained above, the relative complexity of English sentences can be quantified using 3ECM (CMM, BOM and SCM) mechanisms of word order change and movement of constituents that cause major grammatical and pattern changes . Also, 3ECM represented by combination functions, i . e . , the sentence complexity functions f{a , β ) = [ α CMM] / [ β BOM] and f(α , β , Y ) = [Y SCM ( [ α CMM/β BOM] ) ] , can be used to classify the development stages of English sentences and formulate the critical structures governed by the ECM combination of each development stage .

The English development stages and the critical structures of each stage are summarized in Table 4. [Table 4]

Stages ECM Functions

Aa , β ) = [α CMM]/[β BOM] and Λα , β , γ ) = = [γ SCMC [α CMM/β BOM] )J

ECM : Simple Sentence/Main Clause (Semi- ECM : Embedded Clause Complex)

[α CMM]/[ β BOM] L'SCM([α CMM]/[β BOM] )] [Y SCM( [α CMMM β BOM] )]

Simple IX [^•/+] •

Sentence 2X [+/+]

3X [+/-]

4X [-/-I

Semi— 5X a [^•([^•/+])]

As shown in Table 4, English sentences are classified into simple, semi-complex, complex and multiple-complex sentences according to the complexity functions showing the relative complexity of the simple, semi-complex and complex sentences . As the lowest stage, IX is the starting point of English development stages . NX"^"1 is the highest development stage. According to ECM analysis, every sentence in English is classified into one of the development stages from IX to nX^11"1. The basic principle of the ECM combination of each development stage in Table 4 will be discussed in further detail.

First, development stages IX to 4X of simple sentences have

ECM combination values [ ^• /+] , [+/+] , [+/-] and [-/-] , respectively, according to the hierarchical combinations of the + or - value of variables α and jβ in the ECM combination

[ α CMM/β BOM] . The ECM combination values constitute an implicational scale, i . e . , [ ^• /+] C [+/+] C [+/-] C [-/-] . ("C" means "implicational relation. " For example, ACB means that B includes A but A does not include B. ) As is clear from Table 4, the development stages of simple sentences have implicational relations of 1XC2XC3XC4X. In the development stages of English sentences, an upper stage sentence can include an ECM combination of a lower stage sentence and a critical structure thereof according to the implicational scale . Conversely, a lower stage sentence cannot include an ECM combination of a higher stage sentence . The hierarchical combinations [ ^• /+] , [+/+] , [+/-] and [-/-] of variables α and β in [ α CMM/β BOM] are also applied to semi-complex sentences and embedded clauses of complex sentences . Therefore, the variable values in the ECM combinations of semi- complex and complex sentences are also in a hierarchical construction. In other words, simple sentences, semi-complex sentences and simple sentences all constitute an implicational scale in which any lower stage is assumed to be a prerequisite for any higher stage . The implicational scale implies that each development stage of English sentences is a gradual shedding of ECM combinations . Like a butterfly that grows from a caterpillar through shedding processes, each development stage of English sentences is reached after shedding of the ECM combination of a previous stage . In other words, the ECM combination of each development stage should be shed along a predetermined development route to reach the next stage . English learners can learn English of a higher development stage only when they master the ECM of the lower development stages . A caterpillar cannot do any later stage shedding first or skip any shedding stage to become a butterfly. Similarly, English learners cannot skip any development stage or reverse the development stages in English language acquisition. The implicational scale will be explained in further detail with Example 16.

[Example 16]

(a) *W2iere [John kissed Mary] _ last night? ( [+CMM] / [+BOM] : 2X) (b) [Did John kiss_ Mary] at the part last night? ( [+CMM] / [-B0M] : 3X)

(c) [Where did John kiss_ Mary] _ last night? ( [-CMM] / [-BOM] : 4X)

Sentence (a) has a Wh-fronting structure of development stage 2X which is defined by [+CMM] / [+B0M] of ECM. Sentence (b) has a Y/N-question structure of development stage 3X which is defined by [+CMM] / [-B0M] of ECM. Sentence (c) has a Wh-question structure of development stage 4X which is defined by [-CMM] / [- BOM] of ECM. The scale values of variables α and β in [ α CMM/ jβ BOM] are "+/+, " "+/-" and "-/-" which constitute an implication scale . Accordingly, the ECM of sentence (c) includes the ECM of sentences (a) and (b) and the critical structures thereof . The ECM of sentence (b) includes the ECM of sentence (a) and the critical structures thereof . Conversely, the ECM of sentence (a) cannot include the ECM of sentence (b) and the ECM of sentence (b) cannot include the ECM of sentence (c) in light of the implicational relations . Only those who can compose a sentence in [SVO] pattern ( [ ^• /+] ) can move a constituent in the sentence ( [+/+] ) and then change the basic order ( [+/-] ) . Also, only those who can move the basic order ( [+/-] ) can make a double movement of constituents ( [-/-] ) .

The learnability/teachability principles applicable to English teaching, learning and evaluation can be established from the implicational relations . According to the principles, learners can most effectively learn or acquire language of a specific development stage when they are at least in the proficiency level corresponding to the directly preceding development stage .

Referring to Example 16, when a user wishes to learn English sentences of stage 4X (like sentence (c) ) , he or she should be able to understand the ECM of preceding stage 3X ( [- BOM] ) and create sentences of stage 3X (like sentence (b) ) . If the user' s English proficiency level is stage 2X (sentence (a) ) , the user cannot effectively learn English sentences of stage 4X. The same can be said to those who study mathematics . Only those who know how to add and subtract can learn how to multiply and divide . The learnability /teachability principles are also applicable to error corrections in English. If a user creates a sentence with a grammatical error like sentence (a) in Example 16, the error is deemed to be a systematic error naturally made under [+CMM] / [+BOM] of the ECM at development stage 2X. Since only those who can apply [-CMM] / [-BOM] of the ECM at development stage 4X to English sentences can correct the grammatical error in sentence (a) , it is meaningless or in vain to explain to the user having the English proficiency level corresponding to development stage 2X how the grammatical error (like sentence (a) ) should be corrected. Likewise, it is fruitless to try to tell students who cannot even multiply how to correct errors in divisions .

A method for development stage classification of the sentences in answer data has been explained above . After the sentence development stage classification, a data value corresponding to the development stage is transferred to the evaluation system 200 where the user' s English ability will be evaluated. The evaluation system 200 evaluates the user' s English ability based on the data value received from the sentence development stage classifier system 100. The evaluation process will be explained in detail with reference to FIG. 22.

FIG. 22 is a flow chart showing a process for evaluating English ability according to the present invention. The process comprises the steps of: storing a data value received from the sentence development stage classifier system 100, which corresponds to the development stage of each sentence in the answer data, in the storage unit 210 (S301) ; extracting the data value stored in the storage unit 210 (S302) ; evaluating the extracted data value to calculate a correct answering rate in the answer data (S303) ; comparing the calculated correct answering rate with a reference correct answering rate previously set for passing the stage (S304 ) ; and determining a corresponding English ability level of the user who inputted the answer data (S305) .

If five questions are offered in evaluation data of a specific stage, the correct answering rate can tell how many answers inputted by the user are correct . As a baseline value, the reference rate (%) is a preset baseline or reference point at or above which users should obtain in answer data to pass a specific stage of English ability evaluation. The same reference rate (for example, 100%) of correct answering may be set for every stage of evaluation data . Alternatively, different reference rates may be set for different stages of evaluation data- For example, 100% can be set as the reference correct answering rate for passing stage IX of evaluation (lowest development stage of simple sentences) , 70% for passing stage 5Xc of evaluation (middle development stage of semi-complex sentences) and 50% for passing stage 5X² of evaluation (highest development stage of complex sentences) . If the user has obtained a higher correct answering rate than the reference rate 70% preset for stage 5Xc, he or she is recognized as having an English proficiency level of at least stage 5Xc. If the correct answering rate in the answer data is higher than the reference rates preset for stages IX to 4X (simple sentences) but lower than the reference rate preset for stage 5Xa (complex sentences) , the user is recognized as having a English proficiency level corresponding to stage 4X.

Upon evaluation of the user' s English ability through the above process, the diagnosis/prescription system 300 extracts diagnosis data and prescription data corresponding to the evaluation results from the data storage unit 400 and outputs the extracted data through the text output means 260 and the speech output means 270. The diagnosis data tells which stage the user' s current English proficiency level is . The prescription data gives instructions (how to study) to improve the English ability. For example, in a sentence construction ability evaluation, diagnosis data telling "You can construct sentences of up to stage * but still lack an ability to construct sentences of complexity and difficulty level corresponding to stage ** " can be outputted in text or speech through the text output means 260 or the speech output means 270.

The English ability evaluation, diagnosis and prescription method according to the present invention further includes the step of providing the user with an opportunity to study English using evaluation data-to-be-learned (learning data) which corresponds to the user' s English ability, after completion of the diagnosis and prescription.

Like the evaluation data, the learning data consists of three kinds of data, i . e . , sentence construction ability acquisition data, sentence translation ability acquisition data and dialogue ability acquisition data . The learning data is outputted through the text output means 260 or the speech output means 270. Any answer data in response to the learning data is inputted through the speech input means 280 or the text input means 290. The three kinds of learning data can be outputted automatically according to programs . Alternatively, the user can select any desired data to be outputted.

FIG. 23 is a flow chart showing a process for learning English according to the present invention . The process includes the steps of : outputting evaluation data-to-be-learned which meets the user' s English ability (evaluated) (S401 ) ; receiving answer data inputted in response to the evaluation data-to-be- learned (S402 ) ; transferring the answer data to the sentence development stage classifier system 100 (S403) ; receiving a data value corresponding to the development stage of each sentence in the answer data from the sentence development stage classifier system 100 ; storing the data value (S405) ; evaluating the data value to calculate a correct answering rate in the answer data (S406) ; comparing the calculated correct answering rate with a reference rate preset for passing the stage (S407 ) ; and determining whether the user' s English ability level has changed (S408 ) . The process may further include the step of outputting higher stage evaluation data-to-be-learned if it is determined that the user' s English ability level is improved (S409) .

At step 409, higher stage evaluation data-to-be-learned which meets the user' s current English ability level can be outputted upon the user' s demand or automatically after the determination step 408. At step 401, all evaluation data-to-be- learned can be outputted sequentially by stages (from IX to X³) . Alternatively, evaluation data-to-be-learned can be outputted from a specific stage according to the results of evaluation. For example, if the user' s English ability has been evaluated to be a level corresponding to stage 5Xc (semi-complex sentences) , it will be preferable to output evaluation data-to-be-learned from stage 5Xd.

FIGs . 24 and 25 are tables showing test results obtained by a plurality of test applicants using an English ability evaluation/diagnosis/prescription system and method according to the present invention. In the test, five evaluation data were provided at each stage . When an applicant obtained 80% or a higher correct answering rate in the answer data inputted in response to five evaluation data of a specific stage, he or she is deemed to have passed the stage . When two or more of the five answers are wrong, the applicant is deemed to have failed to pass the stage.

In the tables of FIGs . 24 and 25, the leftmost columns represent development stages . At the bottom of the table, "si, s2, s3, ..." are serial numbers assigned to the test applicants and "TC" represents TOEIC scores of the test applicants . The TOEIC scores were inserted in the table for purpose of comparison with evaluation results obtained by the applicants using the system and method according to the present invention.

Also, "DS" at the bottom of the table refers to a development stage corresponding each applicant' s English ability level evaluated according to the present invention. For example, "applicant s2, stage 5X²" in the table means that the applicant s2 has passed all evaluation of stages from IX to 5X². In the table, "+" indicates a right answer, "-" a wrong answer and "=" an almost right or similar answer.

As shown in FIGs . 24 and 25, a significant correlation was found between the development stages assigned to the applicants si to s33 according on their results of an English sentence construction test and the TOEIC scores of those applicants . Applicants assigned high development stages as a result of the English sentence construction test have relatively higher TOEIC scores . Conversely, applicant assigned low development stages have relatively lower TOEIC scores .

However, there exists no significant correlation between the development stages assigned to the applicants s34 to s38 in the English sentence construction test and the scores they obtained in TOEIC. In other words, certain applicants having high TOEIC scores were assigned relatively lower development stages in the English sentence construction test . Such discrepancy arises not because the English ability evaluation/diagnosis/prescription system of the present invention is problematic or unreliable, but because TOEIC scores are not a reliable indicator of productive English skills . In a free English writing test, most of the applicants having high TOEIC scores could combine simple sentences merely using "and, " "so" or "but . " They apparently lacked skills of composing complex sentences using a connective, conjunction or the like .

Although a preferred embodiment of the present invention has been described for illustrative purposes, those skilled in the art will appreciate that various modifications, additions and substitutions are possible, without departing from the scope and spirit of the invention as disclosed in the accompanying claims .

Claims

WHAT IS CLAIMED IS :

1. An English ability evaluation/diagnosis/prescription system having sub-systems including: an evaluation system for outputting evaluation data for English ability evaluation and evaluating a user' s English ability based on answer data inputted by the user; a sentence development stage classifier system for classifying sentences included in the answer data into a pertinent development stage and transferring results of classification to the evaluation system; a diagnosis/prescription system for diagnosing the user' s English ability based on results of evaluation outputted from the evaluation system and providing the user with an individualized prescription; a data storage unit for storing data necessary for the English ability evaluation, diagnosis and prescription; and a control unit for controlling overall operations of the sub-systems to provide accurate evaluation, diagnosis and prescription concerning the user' s English ability.

2. The system as claimed in claim 1, wherein said sentence development stage classifier system analyzes the structural complexity and difficulty level of the sentences in the answer data inputted by the user in response to the evaluation data and classifies the sentences in the answer data into a pertinent development stage based on the analysis .

3. The system as claimed in claim 1, wherein said diagnosis/prescription system includes a learning system for outputting evaluation data-to-be-learned which corresponds to the evaluation results received from the evaluation system, evaluating answer data inputted in response to the evaluation data-to-be-learned and adjusting the difficulty level of the data-to-be-learned to allow the user to keep learning.

4. The system as claimed in claim 2, wherein said sentence development stage classifier system includes : a constituent movement detector for classifying the answer data by sentence pattern (simple, semi-complex or complex) , detecting any grammatical structure change in each sentence of the answer data as compared to a standard grammatical structure of each sentence pattern and outputting a first variable value ( α ) corresponding to the detected grammatical structure change; a word order change detector for detecting any change of word order in each sentence of the answer data as compared to a standard word order in each sentence pattern and outputting a second variable value ( JS ) corresponding to the detected word order change; and a sentence combination detector for detecting any grammatical structure change in an embedded clause of a complex sentence included in the answer data and outputting a third variable value ( Y ) corresponding to the detected grammatical structure change .

5. The system as claimed in claim 4, wherein said sentence development stage classifier system classifies each sentence in the answer data into a pertinent development stage using a simple sentence complexity function defined by the first and second variable values when the sentence in the answer data is a simple sentence .

β . The system as claimed in claim 4, wherein said sentence development stage classifier system classifies each sentence in the answer data into a pertinent development stage using a semi-complex sentence complexity function defined by the first and second variable values when the sentence in the answer data is a semi-complex sentence .

7. The system as claimed in claim 4, wherein said sentence development stage classifier system classifies each sentence in the answer data into a pertinent complex sentence development stage, if the sentence is in a complex pattern composed of at least two sentences acting as a main clause and an embedded clause, by classifying the main clause using a simple or semi-complex sentence complexity function defined by the first and second variable values and the embedded clause using an embedded clause complexity function defined by the first, second and third variable values and combining the development stages of the main clause and the embedded clause together.

8. The system as claimed in claim 4 , wherein said constituent movement detector compares each sentence in the answer data with a standard grammatical structure of the sentence pattern to detect any movement of constituent and outputs the first variable value depending on whether only one constituent is moved (single movement) or two constituents are moved (double movement) .

9. The system as claimed in claim 4 , wherein said word order change detector compares each sentence in the answer data with a basic order of the sentence pattern to detect any change of word order and outputs the second variable value depending on whether the basic word order has changed or not .

10. The system as claimed in claim 4, wherein said sentence combination detector compares an embedded clause of a complex sentence in the answer data with a standard embedded clause structure to determine whether a constituent of the embedded clause is moved within or out from the embedded clause and outputs the third variable value depending on the determination.

11. The system as claimed in any of claims 1, 2 and 4 to 7, wherein said sentence development stage classifier system classifies a plurality of answer data inputted by the user in response to a plurality of evaluation data of each stage into a pertinent development stage and transfers the results of classification to the evaluation system.

12. The system as claimed in claim 11, wherein said evaluation system includes : a storage unit for storing a value inputted from the sentence development stage classifier system which corresponds to the development stage of each sentence; an extractor/output unit for extracting the development stage value stored in the storage unit and assessing the extracted value to produce a correct answering rate in the answer data inputted by the user in response to the evaluation data of a specific stage; and a determination unit for determining the user' s English proficiency level based on the correct answering rate .

13. The system as claimed in claim 12, wherein said evaluation system further includes : a text output means for outputting the evaluation data in text; a speech output means for outputting the answer data in speech; a speech input means for inputting the answer data in speech; and a text input means for inputting the answer data in text .

14. The system as claimed in claim 13, wherein said answer data inputted through the speech input means or the text input means is transferred to the sentence development stage classifier system.

15. The system as claimed in claim 3, wherein said learning system outputs the evaluation data-to-be-learned to the evaluation system, evaluates the answer data inputted in response to the evaluation data-to-be-learned based on a value received from the sentence development stage classifier system which corresponds to the development stage of the answer data and determines whether to keep providing data-to-be-learned or adjust the difficulty level of the data-to-be-learned.

16. The system as claimed in claim 1, wherein said data storage unit consists of : a first data storage unit for storing evaluation data in text; a second data storage unit for storing evaluation data in speech; a third data storage unit for storing diagnosis /prescription data corresponding to each stage; and a fourth data storage unit for storing reference rates of correct answering for passing the respective stages of evaluation .

17. The . system as claimed in claim 16, wherein said evaluation data consists of question data and example sentence data classified according to the structural complexity and difficulty levels of sentences .

18. The system as claimed in claim 17 , wherein said evaluation data includes sentence construction ability evaluation data, translation ability evaluation data and dialogue ability evaluation data .

19. A method for evaluating a user' s English ability and providing the user with an individualized diagnosis and prescription to improve the English ability, which comprises the steps of: outputting evaluation data classified by stage according structural complexity and difficulty levels of English sentences; analyzing the structural complexity and difficulty level of sentences in answer data inputted by the user in response to the evaluation data to classify each sentence in the answer data into a pertinent development stage; calculating a correct answering rate in the answer data and evaluating the user' s English ability based on the correct answering rate; and providing the user with an individualized diagnosis and prescription based on results of evaluation .

20. The method as claimed in claim 19, wherein said evaluation data is outputted sequentially by stages classified according to the structural complexity and difficulty levels of sentences or randomly regardless of the stages .

21. The method as claimed in claim 19, wherein said evaluation data is differently outputted in an English sentence construction test, English sentence translation test and English dialogue test .

22. The method as claimed in claim 21, wherein one of said tests is selected automatically or upon the user' s selection .

23. The method as claimed in claim 19, wherein said evaluation data consists of example sentences and questions outputted in text or speech.

24. The method as claimed in claim 19, wherein said evaluation data is outputted sequentially by stages classified according to the structural complexity and difficulty levels of sentences .

25. The method as claimed in claim 23, wherein said example sentences and questions are sequentially outputted in predetermined time intervals .

26. The method as claimed in claim 19, wherein said answer data is inputted in text or speech.

27. The method as claimed in claim 25, wherein said example sentences and questions are each displayed in a flash form that disappears after a few seconds .

28. The method as claimed in claim 19, wherein said step of classifying each sentence in the answer data into a pertinent development stage includes : recognizing a pattern of the sentence; detecting any change of grammatical structure and word order in the sentence in view of standard grammatical structure and word order of the pattern; and classifying the sentence into a pertinent development stage based on variable values assigned according to the detected change in grammatical structure and word order.

29. The method as claimed in claim 28 , wherein each sentence in the answer data, if it is a simple sentence, is classified into a pertinent development stage based on the first variable value corresponding to a grammatical change detected in the sentence in view of a standard grammatical structure and the second variable value corresponding to a word order change detected in the sentence in view of a basic simple sentence word order.

30. The method as claimed in claim 29, wherein said sentence in the answer data is classified into : a first development stage when the sentence has neither a movement of constituent nor a change of word order in view of standard simple sentence structure and word order; a second development stage when the sentence has no word order change despite a constituent moved or added thereto; a third development stage when the sentence has a word order change due to a new constituent added or moved to the sentence; or a fourth development stage when the sentence has a double movement of constituents which results in a change of word order.

31. The method as claimed in claim 28, wherein each sentence in the answer data, if it is a semi-complex sentence, is classified into a pertinent development stage using the same method as used in simple sentence classification.

32. The method as claimed in claim 28 , wherein each sentence in the answer data, if it is a complex sentence composed of at least two sentences acting as a main clause and an embedded clause, is classified into a pertinent complex sentence development stage by classifying the embedded clause based on the first variable value corresponding to a grammatical change detected in the embedded clause, the second variable value corresponding to a word order change detected in the embedded clause and the third variable value corresponding to a movement of constituent within or out from the embedded clause .

33. The method as claimed in claim 32, wherein said main clause is classified using the same method as used in simple sentence classification if it is a simple sentence or using the same method as used in semi-complex sentence classification if it is a semi-complex sentence .

34. The method as claimed in claim 32 , wherein said complex sentence in the answer data is classified into : a first development stage when the combination of the main clause and the embedded clause has caused neither a movement of a constituent included in the embedded clause nor a change of word order; a second development stage when the combination of the main clause and the embedded clause has caused a single movement of a constituent originally included in the embedded clause within the embedded clause but no change of word order; a third development stage when the combination of the main clause and the embedded clause has caused a single movement of a constituent originally included in the embedded within the embedded clause and a change of word order; a fourth development stage when the combination of the main clause and the embedded clause has caused at least two movements of constituents originally included in the embedded clause within the embedded clause and a change of word order; or a fifth development stage when the combination of the main clause and the embedded clause has caused at least two movements of constituents originally included in the embedded clause out from the embedded clause and a change of word order.

35. The method as claimed in claim 19, wherein said step of evaluating the user' s English ability includes : storing a data value corresponding to the answer data of each stage; extracting the stored data value of each stage; evaluating the extracted data value to calculate a correct answering rate in the answer data; comparing the calculated correct answering rate with a reference correct answering rate previously set for passing the stage; and determining a corresponding English ability level of the user who inputted the answer data .

36. The method as claimed in claim 35, wherein said determination of the English ability level of the user is made based on the results of comparison between the correct answering rate in the answer data inputted in response to the evaluation data by stage with the reference rate previously set for passing each stage .

37. The method as claimed in claim 36, wherein said determination of the English ability level of the user includes : recognizing the user as having an ability to input answer data of at least a specific stage if the correct answering rate in the answer data is higher than a reference rate previously set for passing the specific stage .

38. The method as claimed in claim 19, further comprising the step of providing evaluation data-to-be-learned which corresponds to the results of diagnosis and prescription to enable the user to keep learning using evaluation data suitable to the user' s English ability level .

39. The method as claimed in claim 38 , wherein said evaluation data-to-be-learned consists of sentence construction ability acquisition data, sentence translation ability acquisition data and dialogue ability acquisition data .

40. The method as claimed in claim 39, wherein any of said three kinds of data is outputted according to the user' s selection.

41. The method as claimed in claim 38 , where said learning step includes : outputting evaluation data-to-be-learned; allowing the user to input answer data in response to the evaluation data-to-be-learned; transferring the answer data to a sentence development stage classifier system; receiving a data value corresponding to the development stage of each sentence in the answer data from the sentence development stage classifier system; storing the data value corresponding to the development stage of each sentence in the answer data; evaluating the data value to calculate a correct answering rate in the answer data; comparing the calculated correct answering rate with a reference correct answering rate previously set for the stage; and determining whether the user' s English ability level has changed.

42. The method as claimed in claim 41, wherein said determination of any change of the user' s English ability level is made based on the results of comparison between the correct answering rate in the answer data with the reference rate previously set for passing the stage .

43. The method as claimed in claim 41, wherein evaluation data-to-be-learned selected by the user is outputted upon the user' s demand or evaluation data-to-be-learned that meets the development stage classified as a result of evaluation is automatically outputted.

44. A computer-readable memory device for recording programs for English ability evaluation, diagnosis and prescription. The programs are used to provide a user with evaluation data in text or speech which is classified by development stage according to the structural complexity and difficulty levels of English sentences, receive answer data inputted by the user in text or speech in response to the evaluation data, analyze the structural complexity and difficulty level of each sentence in the answer data to classify the sentence into a pertinent development stage, evaluate the user' s English ability based on a correct answering rate in the answer data, provide the user with an individualized diagnosis and prescription and, if necessary, output evaluation data-to-be- learned so that the user can keep learning to improve his or her English ability.