Comparison Results of Google Translate and Microsoft Translator on the Novel Mughamarah Zahrah Ma'a Ash-Syajarah by Yacoub Al-Sharouni

: This research aims to compare the translation results of Google Translate and Microsoft Translator based on grammatical aspects devoted to fi'il ma’lum-majhul , zaman al-fi'l , and dhamir . This type of study is qualitative, comparative, and descriptive. The main source of data in this research is the novel Mughamarah Zahrah Ma'a Ash-Shajarah by Yacoub Al-Sharouni. Secondary data sources are literature related to Arabic grammatical rules, google Translate translation, and Microsoft translator. The data collection technique is translating a novel by a sentence from Arabic to Indonesian using Google Translate and Microsoft Translator and then recording it. Data validation techniques are performed by triangulation of data and time. Data analysis techniques use data reduction techniques, data presentation, and conclusions. The result of this research is that Google Translate produces good translations in terms of fi'il ma'lum-majhul. Google Translate and Microsoft Translator are inconsistent in translating zaman al-fi'l , while Microsoft Translator produces good translations in terms of dhamir .

Based on this simple experiment, of course it can be seen that the translation results from Google Translate and Microsoft Translator present two very different results.If Google Translate and Microsoft Translator are used in everyday life unwisely, misunderstandings or miscommunication will arise between communicants.Therefore, it is very appropriate for the statement made by Rokhman and Surahmat that the difference in translation results between Google Translate and Microsoft Translator can have a big impact on all aspects of life, especially in the world of translation (Rokhman & Surahmat, 2020, p. 49).The difference in translation results between Google Translate and Microsoft Translator can have a positive impact because it can make it easier for humans to find out more precise translation results.

Conversely, differences in translation results between Google Translate and Microsoft
Translator can negatively impact if both produce incorrect translations due to grammatical incompatibilities.The number of users of the two translation machines is not without reason, but because of the several advantages that both have.The advantages of Google Translate include being easy and practical to use, free of charge and can be applied through text, voice and camera modes.
The advantages of Microsoft Translator include that it can be used even without an internet connection, is free of charge and can be used through text, camera and voice modes.But behind its advantages, both translation machines have the same weakness, which cannot accurately translate long texts and often grammatical errors occur (Muscutt, 2022).
The data above shows that Google Translate and Microsoft Translator can only function as translation tools and cannot replace the role of humans in the field of translation.According to researchers, this is because translation is a complicated process and requires in-depth analysis related to language selection, grammatical composition and even culture of the source language and target language which the process cannot be done by machine translation.The researcher's statement is in line with the opinion that computers are clearly different from humans because computers or machines can only process according to the input they receive, while humans have innate abilities in language and logic and expression that can develop according to their experience (Bolander, 2019, p. 850) Based on translation rules, a good translation is one that can be accepted and understood by readers.To be accepted and understood by the reader, it is necessary to know the rules of the source language and the target language, especially in grammatical equivalence in the form of dhamir, zaman al-fi'l, and fi'il ma'lum-majhul in order to produce a contextual translation (Uyuni, 2023, p. 61).Dhamir in Arabic is divided into three based on the speaker namely mutakallim (first person pronoun), mukhatab (second person pronoun) and ghaib (third person pronoun).Dhamir is divided into two based on the number and gender aspects namely mudzakkar (Man) and muannats (Women).Dhamir based on the number aspect is divided into three, namely mufrad, tasniyah and jamak.(Switri, 2022).Arabic distinguishes verbs based on their timing, so there is the term fi'il madhi to refer to verbs in the past and fi'l mudhari' to refer to verbs that are being done or to be done (Amin, 2022).In Arabic there is also active voice and passive voice.The active voice in Arabic is indicated by the presence of fi'l (verb), fa'il (subject) and maf'ul (object), while the passive voice consists only of fi'l (verb) and naibul fa'il (object that replaces the subject's position).The passive voice and active voice in Arabic can be seen from the change in harakat (punctuation) in fi'il (verb) (Saomi & Romlah, 2020).
The use of machine translation to facilitate the translation process must still pass the process of double-checking the translation results.A review of the translation results is very necessary to analyze language errors in terms of various aspects, especially in grammatical aspects (Santoso, 2010).Comparing translation results by Google Translate and Microsoft Translator is very necessary in order to find out which translation machine is more accurate.
This makes researchers interested to know which of Google Translate and Microsoft Translator is more accurate in translating grammatical equivalents in the form of dhamir, zaman al-fi'l, and fi'il ma'lum-majhul.
There are several studies that previously examined the quality of translation results such as: gender bias in translation using machine translation (Nurtamin, 2022), morphological errors in machine translation results (Alawi, 2019;Ilmi, 2019;Tawfik et al., 2019), sematic errors translated using Google Translate and Microsoft Translator (Ameur et al., 2019;Brour & Benabbou, 2021;Ding et al., 2019;Fayruza et al., 2020;Gashaw & Shashirekha, 2019;Yaakub et al., 2020) comparison and misanalysis of Google Translate's translation results from Arabic to Indonesian and vice versa (Abidin et al., 2020;Simanjuntak, 2019;Sopian et al., 2021;Untara & Setiawan, 2020), General translation quality (Fayruza et al., 2020;Khoiriyah, 2020)    The translation above has grammatical differences.Translation with google translate converts the source language into the target language with an active sentence pattern where there is a subject "dia" (fa'il) and a verb "duga" (fi'il) but no object in it.This is in accordance with the Arabic rule which states that the sentence arrangement consisting of fi'il and fa'il is an active sentence pattern (Saomi & Romlah, 2020).The target language is Indonesian language in this research state that active sentences are sentences consisting of the subject as the perpetrator and the predicate as an active verb (Bachrudin, 2023).This indicates that Google Translate translation tries to preserve the grammar of the source language but there are particles that are discarded in order to reduce word waste.
The translation with Microsoft Translator converts Arabic grammatical in the form of active voice consisting of verbs (fi'il), subject (fa'il) and object (maf'ul) translated into passive voice in Indonesian.Indonesian language rules state that passive voice is a sentence that makes the subject an object and one of the other signs is that there is an affix "di" at the beginning of the predicate word (Bachrudin, 2023).The result of the Microsoft translator translation has a "tidak diharapkannya" clause.The translation does not mention the subject, but there is an affix "di" in the word "harapkan" so that it can be ascertained that the translation of Microsoft The word َّ ‫اضطر‬ is fi'il madhi mabni ma'lum or in Indonesian is an active voice.The above sentence based on Arabic rules is an active voice because the verb or fi'il َّ ‫اضطر‬ ‫و‬ has a subject or fa'il in the form of ‫أمحد‬ and the object is the ‫الصب‬ ‫لرغبة‬ ‫يستجيب‬ ‫أن‬ .The appropriate translation result is the result of translation using google translate because it still maintains the grammar of the source language, which is still producing active sentence translations in the form of "Ahmad al-Nashar" as the subject, "terpaksa" as a predicate and "menuruti keinginan bocah itu" is an object in the form of a clause.This is in accordance with the rules of the source language (Arabic) related to fi'il mabni ma'lum which consists of fi 'il, fa'il and maf'ul (Saomi & Romlah, 2020).The results of the translation of google translate into the target language (Indonesian) are also in accordance with the rules of Indonesian because it consists of subjects, predicates and objects (Bachrudin, 2023).
In contrast to the results of Microsoft Translator translations that do not maintain the rules of the source language, namely a change from active voice (Bsu) to passive voice (BSa).As explained earlier that the findings of the data mentioned above are fi'il mabni ma'lum or in Indonesian called active voice because in accordance with Arabic rules, namely fi'il mabni ma'lum consists of fi 'il, fa'il and maf'ul (Saomi & Romlah, 2020).Then Microsoft translator changed the grammar to passive voice in Indonesian.The evidence that shows the existence of passive voice in the findings of the above data is "Ahmed Al-Nashar was forced".The predicate "dipaksa" comes from the verb "paksa" which is affixed "di".In addition, the subject of "Ahmed Al-Nashar" became subject to "dipaksa" work.The addition of "di" to the predicate and the position of the subject subject subjected to work is a passive voice pattern in Indonesian (Bachrudin, 2023).The presentation shows that the translation of Microsoft translator changes the grammar of source language from active voice to passive rules in target language.
After comparing the two translation results, researchers argue that the results of Google translate translation are more in accordance with the rules of Indonesian than the results of Microsoft Translator translation.Such a comparison is the same as previous research conducted by (Fayruza et al., 2020) which compares the translation quality of Google Translate and Microsoft Translator.But there is also a difference between this study and the study, namely in the aspect of language.Fayruza et al compared the translation results of the two machines in general, while this study focused more on grammatical aspects.

Zaman al-fi'li
The difference between Google Translate and Microsoft Translator is also found in the form of zaman al-fi'li or verb tenses.Verb tenses relates to the time that marks a job.The Arabic rules that explain verb tenses or zaman al-fi'l are divided into three, namely fi'il madhi to mention verbs in the past and fi'l mudhari' to mention verbs that are being done or to be done (Amin, 2022).The difference between Google Translate and Microsoft Translator translations from the time of al-fi'l can be seen from the sentences below:
2. The translation results using google translate are more in accordance with Arabic rules than the translation results of Microsoft Tranlator.In the translation of google translate there is a clause ‫اهتم‬ ‫ملا‬ ‫و‬ which consists of fi'il madhi ‫.اهتم‬The Arabic rule states that fi'il madhi is a verb that has a past meaning (Switri, 2022).And in Indonesian rule, the meaning of time is explained using adverbs will, past, medium etc (Muam & Nugraha, 2020).Fi'il madhi ‫اهتم‬ becomes "peduli" without mentioning any information or a specific period of time so as to give a past impression.This is in accordance with the Arabic rule that ‫اهتم‬ is fi'il Madhi which indicates the past (Switri, 2022).

‫ملا‬ ‫و‬ ‫ا‬ ‫أحد‬ ‫هتم‬ ‫ا‬ ‫ازدادو‬ ‫أهنم‬
Microsoft translator translates ‫اهتم‬ to "ketika ada yang peduli".This is in contrast to the Arabic rule that should translate fi'l madhi into past tense verbs (Switri, 2022), But in the findings of the data there is the word "ketika" which is a description of the time that is happening.The rule Indonesian states that the description of time can be shown through the adverbs will, past, medium etc (Muam & Nugraha, 2020).The translation contains the word "ketika" which in Indonesian indicates the time that is happening, so it is inversely proportional to the source language.From this presentation, it can be seen that the translation results of Microsoft translator do not match the source language.
‫ا‬ ‫ازدادو‬ ‫أهنم‬ ‫أحسست‬ ‫وقد‬ .Fi'il Madhi is a fi'il or verb that has a past time (Switri, 2022).The Microsoft translator translation contains the word "has" which indicates the time that has passed.This is in accordance with Indonesian rule which states that the information of time can be shown through the adverbs will, past, medium etc (Muam & Nugraha, 2020).
The translation of google translate in that sentence does not mention the time.Sentences that do not mention time do tend to seem past.But the adverb mentioned is more understandable than the word not mentioned.The translation of google translate from the source language ‫ا‬ ‫ازدادو‬ to Indonesian "mereka meningkat" without mentioning the time period can be said to be appropriate, but not quite right.Based on this presentation, the results of Microsoft translator translation are superior to google translate because Microsoft translator produces translations that mention the time clearly.
After comparing the two translation results, researchers argue that the translation results of Microsoft trasnlator are more in accordance with the rules of Indonesian (BSa) than the results of google translate translation in terms of the zaman al-fi'l or verb tenses.Such a comparison is the same as previous research conducted by (Abidin et al., 2020;Simanjuntak, 2019;Sopian et al., 2021;Untara & Setiawan, 2020) Which discusses the comparison and analysis errors of Google Translate translation results from Arabic to Indonesian and vice versa.In the study, there was an analysis error produced by the google translate translation engine.Analysis errors are also found in this study which is shown in Google Translate's analysis errors in translating verb tenses or zaman al-fi'l.

Dhamir
Google Translate and Microsoft are machine translators that can help humans in the translation process, but in their use must be accompanied by the natural ability of translators so that the translation results do not experience errors.The mistake that often occurs when using machine translation is an error in pronouns translation or in Arabic is dhamir (Nurtamin, 2022).Dhamir in Arabic is divided into three based on the speaker namely mutakallim (first person pronoun), mukhatab (second person pronoun) and ghaib (third person pronoun).
Dhamir is divided into two based on the number and gender aspects, namely mudzakkar (Man) and muannats (woman).Dhamir based on the number aspect is divided into three, namely mufrad, mudzakkar and jamak.(Switri, 2022).
The difference between Google Translate and Microsoft Translator translation results in terms of pronouns or dhamir can be seen from the table and explanation below: Google translate's translation of the word ‫أفاجأ‬ is a mistake.The word ‫أفاجأ‬ translates into Indonesian to "dia dikejutkan".The translation is not in accordance with the rules of the source language (Arabic).The word ‫أفاجأ‬ is fi'il mudhari' which contains the letters mudhara'ah hamzah.Arabic rules explain that the subject (Fa'il) can be indicated by the letters mudhara'ah.The letter mudhara'ah hamzah found in fi'il mudhari' indicates the first person pronoun or dhamir mutakallim (Saomi & Romlah, 2020).The first pronoun in Indonesian is "aku" or "saya" (Trianto, 2006).
Based on Arabic rules, the correct translation result is "Saya dikejutkan" produced by Microsoft Translator.The word ‫أفاجأ‬ is fi'il mudhari' which follows wazan ‫فاجأ‬ -‫يفاجأ‬ , later changed to ‫أفاجأ‬ because it contains dhamir ‫أن‬ or mutakallim marked with the letters mudhara'ah hamzah ‫.)أ(‬This is in accordance with the Arabic rule which states that the letter mudhara'a hamzah contained in fi'il mudhari' indicates the first person pronoun or dhamir mutakallim (Saomi & Romlah, 2020), and the rule of Indonesian which is that the first person pronoun in Indonesian is "aku" or "saya" (Trianto, 2006).Based on this explanation, it can be ascertained that the translation of Microsoft translator is the right translation.Google translate dhamir ‫هو‬ into "inilah".If viewed from the rules of source language and target language, then the results of the translation are not correct.In that sentence there is dhamir ‫هو‬ which is the dhamir ghaib in Arabic rules.Dhamir ghaib in Indonesian is a thirdperson pronoun.Third person pronouns generally mean people or things that are far away or absent from a conversation (Saomi & Romlah, 2020).The rule of Indonesian states that thirdperson pronouns use "dia" or "mereka", while the word "itu" is used to replace objects that are distant or not in place (Trianto, 2006).Based on this, it can be ascertained that the translation results of google translate that translates the word ‫هو‬ into "ini" is not in accordance with the rules of the source language or target language.
The translation results of Microsoft translators who translate dhamir ‫هو‬ into it is more precise because it is in accordance with the rules of the source language (Arabic) and the target language (Indonesian).This is because in that sentence there is dhamir ‫هو‬ which is an dhamir mufrad ghaib in Arabic rules.Dhamir mufrad ghaib in Indonesian is a singular and thirdperson pronoun.Third person pronouns generally mean people or things that are far away or absent from a conversation (Saomi & Romlah, 2020).The rule of Indonesian states that thirdperson pronouns use "dia" or "mereka", while the word "itu" is used to replace objects that are distant or not in place (Trianto, 2006).
Based on the data that has been obtained, the researcher presents a summary of the results of the research on the comparison of the results of google translate and microsoft translator in the novel Mughamarah Zahra Ma'a Syajarah by Jacob Al-Sharoni in tabular form.Researchers compare the harmony of the translation results of several grammatical aspects as Active and passive voice, verb tenses and pronouns in the table below.(Trianto, 2006).
A good translation result is a translation that can produce a translation of pronouns according to the source language and target language.In this case, the number, point of view, and far or near.And this result, microsoft translator is more in line with Indonesian language rules (BSa) than the Google Translate translation.Errors in translating persona or dhamir using google translate translation machines have been researched by (Nurtamin, 2022).But the results of this study are different from this study.In the study, researchers focused on analyzing gender bias so that what was produced was gender refraction in the results of Google Translate translation, while in this study, researchers focused on pronouns or dhamir errors according to their division in Arabic rules.
Based on the results of the comparison and analysis described above, the researcher also saw that the translation results of Google Translate and Microsoft Translator had differences in several aspects, namely aspects of passive voice and active voice (fi'il ma'lum and fi'il majhul) and aspects of tenses.verb (zaman al-fi'l).The results of the translation using Google Translate tend to be more precise in translating passive voice and active voice (fi'il ma'lum and fi'il majhul).This is because in Arabic (source language), verbs are divided into two, Translator on grammatical aspects that have not been studied much by previous researchers.
However, this research is only limited to three grammatical aspects, namely pronouns, tenses, and active-passive sentences so that researchers can still develop this research from other grammatical aspects.This research produced a conclusion stating that Google Translate and Microsoft Translator are two translation machines that can help the translation process, but correction or re-reading related to translation results is very necessary.Google trasnslate and Microsoft translator can be used simultaneously to help the translation process, but there are things that must be considered as described in this study.Google trasnslate and microsoft translator are translation machines that can develop so it is expected that more researchers can study related to these two translation machines both from grammatical, semantic, morphological and technological aspects.Studies related to other machine translations are also needed in order to provide insight into machine translation that can produce good translations.

Arab-Indonesia
As a Neural Machine Translation (NMT)-based translation engine, Google Translate and Microsoft Translator are two of the most popular machine translations used by people around the world.Reporting from the page (Content, 2023), Google Translate is the first order of the most widely used translation engines.Based on data launched by Turovsky in 2016 right on the 10th anniversary of Google Translate in April 2016 stated that Google Translate users reached 100 billion users and the number of users continues to increase every year.Microsoft Translator is the second most used translation engine after Google Translate (Turovsky, 2016).Al-Lisan: Jurnal Bahasa (e-Journal), Volume 8, No.2, August 2023 156

( 1 )Findings
analyze the comparison and translation results of Google Translate and Microsoft Translator in terms of pronouns (dhamir); (2) analyze and compare the translation results of Google Translate and Microsoft Translator in terms of tenses (the time of al fi'l) and; (3) analyze and compare the results of Google Translate and Microsoft Translator translations in terms of passive-active kalimaf (fi'il ma'lum-majhul).So that it is expected to be a consideration to use machine translation as needed.B. RESEARCH METHOD The research method used in this research is qualitative and descriptive because the researcher explained the results of a comparison of translation phenomena using Google Translate and Microsoft Translator machine translators which were analyzed and explained in sentence form.In addition, this research used a literal translation method that seeks to preserve the original form of the source language.The main source of data in this research is the novel Mughamarah Zahrah Ma'a Asy-Syajarah by Yacoub Al-Sharouni published by Dar Al-Ma'arif,, Cairo in 2012 with a total of 39 pages.Secondary data sources in this research are books, journals and other literature related to grammatical commensurability and the use of machine translation.Data collection was carried out by translating the novel Mughamarah Zahrah Ma'a Asy-Syajarah by Yacoub Al-Sharouni sentence by sentence using Google Translate then translating the novel Mughamarah Zahrah Ma'a Asy-Syajarah by Yacoub Al-Sharouni sentence by sentence using Microsoft Translator.Looking for sentences related to grammatical commensurability in the translation results is then mapped based on the aspects that are the focus of research, namely comparing the translation results of google translate and microsoft translator in the aspect of fi'il ma'lum-majhul.time of al-fi'l, and dhamir.Data validation techniques are carried out by triangulation of data.Researchers read the data translated of novel Mughamarah Zahrah Ma'a Asy-Shajarah by Yacoub Al-Sharouni repeatedly and adjusted to various sources related to grammatical commensurability to get a more accurate understanding.Data analysis techniques used in this study are data reduction, data presentation and conclusions.Data reduction taken by researchers is by sorting data and then presenting it in the form of a narrative.The last step used by researchers is to make conclusions based on research objectives.This research analyzes and compares the translation results of google translate and microsoft translator in grammatical aspects that focus on active voice and passive voice or in yang peduli dengan kelahiran atau hidup Anda! (Google Translate) Dan ketika ada yang peduli tentang kelahiran atau hidup Anda!(Microsoft  Translator)

(
mereka meningkat (Google Translate) Dan saya merasa bahwa mereka telah meningkat(Microsoft Translator)    The translation of the word ‫ا‬ ‫ازدادو‬ in the sentence above is more appropriate using Microsoft Translator than google translate.The translation of Microsoft Translator can preserve the grammar of the source language and is more in accordance with Arabic rules in terms of the zaman al-fi'li or verb tenses.The word ‫ا‬ ‫ازدادو‬ in Arabic is fi'il madhi ‫ازداد‬ following wazan ‫افتعل‬ -‫يفتعل‬ ‫بظلى‬ ‫يستمتعان‬ ، ‫حتىت‬ ‫جلسا‬ ‫رجلي‬ ‫من‬ ‫ا‬ ً ‫يوم‬ ‫مسعته‬ ‫ما‬ ‫وهو‬( dengar suatu hari dari dua pria yang duduk di bawah saya, menikmati keteduhan saya (Google Translate) Itulah yang saya dengar suatu hari dari dua pria yang duduk, menikmati keteduhan saya (Microsoft Translator) namely active verbs (fi'il ma'lum) which have an object and passive verbs (fi'il majhul) whose object occupies the subject position.In Arabic, the two verbs can be distinguished based on their harokat or context(Saomi & Romlah, 2020).Whereas in Indonesian (Target Language), active passive verbs are usually indicated by the addition of "di" before the passive verb and the absence of "di" before the active verb(Bachrudin, 2023).In this case Google Translate tend to be more precise between Microsoft Translator.When viewed based on the aspect of verb tenses (al-fi'l era), the translation results of Google Translate and Microsoft Translator are inconsistent.Sometimes the translation of verb tenses or al-fi'l era is more appropriate using Google translate and in other cases it is more appropriate to use a Microsoft translator.inconsistent is caused by the absence of a special equivalence between Arabic (source language) and Indonesian (target language).Arabic divides verbs based on time, namely ‫أمر)‬ ‫فعل‬ ‫مضارع،‬ ‫فعل‬ ‫ماضي،‬ ‫(فعل‬ (Switri, 2022) while Indonesian does not divide verbs based on time.Time in Indonesian is more generally expressed by adverbs such as had, will, now and others(Muam & Nugraha, 2020).Therefore, the results of google translate and microsoft translator cannot detect exactly when it is appropriate to use or not use adverbs.D. CONCLUSIONThisresearch concludes that the results of google translate translation are more accurate in translating Arabic text in terms of active sentences and passive sentences or fi'il ma'lummajhul.Microsoft Translator is superior in translating Arabic text which contains lots of pronouns or dhamir.And google translate and microsoft translator are equivalent in translating texts related to verb tenses or al-fi'l era.The results of this scientific research can reveal new findings related to the comparison of the translation results of Google Translate and Microsoft . Penerbit A-Empat.Yaakub, M. B. H., Sismat, M. A. bin H., & Yunos, I. N. H. M. (2020).Analisis Semantik dan Pragmatik Terhadap Terjemahan Mesin Google Arab -Melayu./doi.org/https://unissa.edu.bn/journal/index.php/jall/article/view/345 (

Table 1 . Active And Passive Voice Translation Results (Fi'il Ma'lum-Majhul)
Arabic, namely Fi'il ma'mul-majhul; Verb tenses or in Arabic called zaman al-fi'li and pronouns or in Arabic is dhamir.Here are some examples of comparisons of google translate and microsoft translator translations in terms of grammatical aspects of fi'il ma'lum-majhul, of novel Mughamarah Zahrah Ma'a Ash-Syajarah by Yacoub Al-Sharouni using Google Translate and Microsoft Translator.Further explanation related to this is as follows: translator changes the grammatical arrangement of Arabic with active voice patterns into passive voice in Indonesian.