English Learning Assistant Application with a Translation Approach

Using Rule Based System

Daniel M. D. U. Kasse, Patrisius Seran and Christa E. B. Bire

Politeknik Negeri Kupang, Penfui, Kupang, NTT, Indonesia

Keywords: English Learning, Dictionary, Rule Based, Translation.

Abstract: English has moved to improve its linguistic status, which is starting to act as a second language and

language

of instruction. This is often found in the social sphere, especially in youth, and education. The use

of English also has an impact on education and the economy. In the field of education, the availability of

sophisticated

translation tools online opens the way for students to translate a text more easily. However, in

real conditions, Indonesian students still find it difficult to practice their English skills. The purpose of this

research is to make

an application as a tool to learn how an English sentence can be translated into

Indonesian. This application not only displays the translation results but also displays the types of words,

sentence patterns to produce

English translations and materials.

1 INTRODUCTION

The use of English in Indonesia in general has a great

influence, although it is still a foreign language.

English has moved to improve its linguistic status,

namely starting to act as a second language and

language of instruction. This is often found in the

social sphere, especially youth, and education. The

use of English also has an impact on education and

the economy.

In the field of education, the availability of

sophisticated translation tools on-line opens the way

for students to translate a text more easily, media that

expose them to English such as programs on

television, music, commands on social media

applications and their gadgets. can increase their

knowledge. on how to understand English text and

convert information into Indonesian. However,

problems can also arise from tools and media that

help them understand English, such as if students

receive translations that do not match the true

meaning after they enter sentences in the source text,

without considering polysemy, transliteration, and

culture. the use of language, the idioms used and so

on, the result of the translation may not be accepted

because it does not convey the idea that the author of

the source text is trying to convey.

English teaching in Indonesia must be improved

if you want to get better results. Student needs

should be the focus of attention in teaching. The

success of teaching English depends on student

achievement in

terms of the goals that have been

determined before

the program starts. English

teachers must be open- minded and ready to keep

learning and strive for better teaching outcomes.

All the necessary information and knowledge must

be used to make

teaching English successful.

English teachers should

be able to choose teaching

materials such as books, journals, audio-video tape

recorders and cassettes, independent access and

computerized language

teaching to facilitate

language learning so that

students can achieve

effective language learning (Richards, 2001:230).

We need to be aware that

language teaching

methods can change over time as can fashion. But

we can always judge whether a

particular method

is suitable for our purposes. It is

important for us

to be open-minded and ready to try new methods to

improve the quality of teaching English. In real

conditions, Indonesian students still

find it

difficult to practice their English language

skills

in communicating. The purpose of this research is to

create an application that not only displays the

results of the translation but also how an English

sentence is translated into Indonesian.

Kasse, D., Seran, P. and Bire, C.

English Learning Assistant Application with a Translation Approach Using Rule Based System.

DOI: 10.5220/0012064600003575

In Proceedings of the 5th International Conference on Applied Science and Technology on Engineering Science (iCAST-ES 2022), pages 1079-1083

ISBN: 978-989-758-619-4; ISSN: 2975-8246

1079

RULE BASED SYSTEM

Rule Based System (RBS) is an expert system that

uses rules to present knowledge. This RBS theory

uses a simple technique, starting with the basic rules

that contain all the knowledge of the problems

encountered which are then coded into if-then rules

containing data, statements and initial information.

The system will check all the if condition rules that

define the subset, conflict sets that exist. If found,

then the system will perform a then condition. This

loop will continue until one or two conditions are met,

if the rules are not found then the system must exit the

loop (terminate).

In the application of the Rule Based System in

the

process of translating sentences from English to

Indonesian, it is based on the rules of English

grammar. English has several sentence patterns,

namely:

•

Pattern 1: Simple Sentence

One independent clause (SV.). E xample:

Mr. Potato Head eats monkeys.

I refuse.

•

Pattern 2: Compound Sentence

Two or more independent clauses. Example:

Mr. Potato Head eats them for breakfast every

day, but I don’t see the attraction.

•

Pattern 3: Complex Sentence

One independent clause PLUS one or more

dependent clauses. Examples: He recommends

them highly because they taste like chicken

when they are hot.

•

Pattern 4: Compound-Complex Sentence

Two or more independent clauses PLUS one

or more dependent clauses. Example: Mr.

Potato Head said that he would share the secret

recipe; however, if he does, Mrs. Potato Head

will feed him to the piranhas, so we are both

safer and happier if I don’t eat monkeys

or steal recipes.

To be able to read input from the user, a

parsing

process with context free grammar is

used. Most

systems for modeling constituent

structures in

English or other natural languages are

using Context

Free Grammar or CFG.

Figure 1: Rules-based translator system.

For a Context Free Grammar has four

parameters (technically called 4-tuple)

•

N: A collection of non-terminal symbols (or

variables){NP, VP, PP}

•

∑: Terminal symbol set {det, noun, verb,..}

•

P: The production set, which is expressed in

the form A → α where A is a non-terminal

symbol {<NP> → det noun}

•

4. S: Start symbol

The parsing method used is top down. The step

is to

find all the appropriate grammar rules. In the

top- down parser there is a strategy called depth-

first

which looks for the appropriate grammar for

each of the first and subsequent entries.

RESULTS AND DISCUSSION

3.1 System Implementation

The translator system is implemented using four

modules that have different functions. Here are the

four modules along with a brief description.

3.1.1 Modul Indentifikasi

The initial stage is the preprocessing stage which is

carried out in this module. At the initial stage, a

checking process will be carried out on user input,

whether it is a word or not (in this case it is a phrase

or sentence). Input in the form of words will be

translated on the translator module while input in the

form of phrases or sentences will undergo a

preprocessing process. The preprocessing stage

includes changing abbreviations (contractions) such

as the word I′ll to be changed to I will. In the

simple future tense, there is the use of tobe going to

which is

changed to will. The tokenization stage is

carried out to get tokens in the form of words

Rules Collection

English

sentence

System Result

Dictionary

iCAST-ES 2022 - International Conference on Applied Science and Technology on Engineering Science

1080

which are then converted into a single form such as

the word kicks which is converted into the word

kick. The next stage

is checking the words and

tenses. Input that meets the

criteria (in this case, all

words are in the word dictionary database and the

tenses used include

simple present tense and simple

future tense) will be

given to the parser module to

perform the scanning process (parsing). The initial

stage is the

preprocessing stage which is carried

out in this

module. At the initial stage, a checking

process will

be carried out on user input, whether it

is a word or not (in this case it is a phrase or

sentence). Input in

the form of words will be

translated on the translator

module while input in

the form of phrases or sentences will undergo a

preprocessing process. The preprocessing stage

includes changing abbreviations (contractions) such

as the word I′ll to be changed to I will. In the

simple future tense, there is the use of tobe going

to which is changed to will. The

tokenization

stage is carried out to get tokens in the

form of

words which are then converted into a single

form

such as the word kicks which is converted into the

word kick. The next stage is checking the words

and tenses. Input that meets the criteria (in this case,

all words are in the word dictionary database and the

tenses used include simple present tense and simple

future tense) will be given to the parser module to

perform the scanning process (parsing).

3.1.2 Modul Parser

This module implements syntax rules in context free

grammar. The process for analyzing syntax rules is

called parsing. The parsing method used is top down

with a depth first strategy (top down depth first

parser). The parser will analyze the syntax structure

of the input provided by the user so that the structure

of the user input is obtained.

Table 1: Syntax rules.

Non terminal Terminal

<NP> → det noun

<NP> → pron

<VP> → verb

The symbol to the left of the arrow is a non-

terminal

symbol and to the right of the arrow is a

terminal

symbol. Table 2 is a description of some of

the

symbols used.

Table 2: Symbol Description.

Simbol Arti

<S> Sentence

<NP> Noun Phrase

<VP> Verb Phrase

adj adjective

pron pronoun

aux auxiliary

det determinant

not negative word

noun noun

verb verb

wh wh-question

propnoun proper noun

inf to infinitive

Pada proses parsing, sistem akan melakukan

pencocokan seluruh aturan sintaks terhadap input

user. Berikut merupakan contoh dari proses parsing

(dengan konsep parsing tree) dalam context free

grammar.

Figure 2: Parsing tree.

From the results of the parsing process, the

input

structure from the user is pron verb det noun.

In the

parsing process there are several processes

English Learning Assistant Application with a Translation Approach Using Rule Based System

1081

including 'expand', 'match', and 'backtrack'. Expand

is the

process of replacing non-terminal symbols

into terminal symbols according to the existing

production rules in the syntax rules such as <NP> to

pron. Match

is the process of matching user input

word types with

syntax rules. Backtrack is the

process of going back to a non-terminal symbol

from a previous production.

3.1.3 Modul Translator

In this module, the MD-DM pattern rules are

implemented, namely the Explaining Explained word

pattern which is commonly found in English texts is

changed to the Explained - Explained word pattern

which is commonly found in Indonesian texts. An

example of the MD pattern is the big house. Table 3

shows the MD-DM pattern rules used.

Table 3: Aturan Pola MD-DM.

No. JK1 JK2 JK_1 JK_2

1 Aux Not Not Aux

2 Adj Noun Noun Adj

3 Det Noun Noun Det

4 Propnoun Noun Noun Propnoun

5 Noun Noun Noun Noun

For user input in the form of word units, the

translation will be carried out directly without going

through these stages. The results of this module are in

the form of word tokens in Indonesian.

3.1.4 Modul Materi

At this stage, the system will display the material

according to the type of words obtained from the

identification.

3.2 System Testing

In the process of testing the system with a black box

can be determined by studying the input and output.

In this test, the focus will be on testing from the

original language (English) to the target language

(Balinese) whether it is in accordance with the

expected results based on the system design and the

suitability of the system interface. Table 4 is some of

the data used in testing the system using the black box

method.

Table 4: Black Box Method Test Data.

Data

Sample

Keterangan Task Status

1 I Love

You

Correct

input

(simple

sentence

pattern)

Process

and

display the

translation

results in

Indonesian

I was

happy

Correct

input

(simple

sentence

pattern)

Process

and

display the

translation

results in

Indonesian

My elder

brother

became

engineer

1988

Input that

has

complex

sentence

pattern

Processes

and

displays a

warning

that Input

has a

complex

sentence

pattern

was happy

Input that

has

unrecognized

pattern

Processes

and

displays a

warning

that the

Input has

pattern not

recognized

CONCLUSION

From the implementation of the concept of translating

English text to Indonesian text using the Rule Based

method as well as from the results of translation

testing as above, several conclusions can be drawn,

namely:

•

This English to Indonesian text translator can

translate sentences in

"simple sentence

pattern" quite well.

•

This English to Indonesian text

translator is

able to recognize the types

of words in the

entered sentence.

•

This research is still in progress so it is

not

yet fully completed

iCAST-ES 2022 - International Conference on Applied Science and Technology on Engineering Science

1082

REFERENCES

E. Gultom, “English Language Teaching Problems in

Indonesia,” 7th Int. Semin. Reg. Educ., vol. 3, (2015).

J. Sudrajat and T. H. Pudjiantoro, “Aplikasi Penterjemah

Bahasa Indonesia Ke

Bahasa Inggris Dengan

Menggunakan Metode Terselia (Supervised

Learning),”

J. Comput. BISNIS, vol. 2, no. 1, (2008).

M. Nadhianti, “An Analysis of Accuracy Level of Google

Translate In English- Bahasa Indonesia and Bahasa

Indonesia-English Translations,” Sastra Inggris-Quill,

vol. 5, no. 4, (2016).

R. A. Sandra, “From English to Indonesia: Translation

Problems and Strategies of EFL Student Teachers - A

Literature Review,” Int. J. Lang. Teach. Educ., vol. 2,

no. 1, (2018), doi: 10.22437/ijolte.v2i1.4520.

S. Simaremare, D. E. Silalahi, P. Saut Raja Sihombing, and

Y. Kristina Sinaga, “The Accuracy of the Translation of

English Verb Phrase into Indonesian Using Google

Translate,” JETAFL Publ., vol. 7, no. 1, (2021).

S. Wahyuningsih and M. Afandi, “Investigating English

speaking problems: Implications for speaking

curriculum development in Indonesia,” Eur. J. Educ.

Res., vol. 9, no. 3, (2020), doi: 10.12973/EU-

JER.9.3.967.

Y. Farahsani, I. P. Rini, and P. H. Jaya, “Google Translate

Accuracy in Translating Specialized Language From

English to Bahasa Indonesia”, Jan. (2021), doi:

10.2991/assehr.k.210120.156.

English Learning Assistant Application with a Translation Approach Using Rule Based System

1083