Free online instruments for text creation and text analysis

202 FREE ONLINE INSTRUMENTS FOR TEXT CREATION AND TEXT ANALYSIS 1 ZHIVKA ILIEVA, PETRANKA IVANOVA, IVELIN IVANOV ИНСТРУМЕНТИ СЪС СВОБОДЕН ДОСТЪП ЗА СЪЗДАВАНЕ И ЗА НАЛИЗ НА ТЕКСТ ЖИВКА ИЛИЕВА, ПЕТРАНКА ИВАНОВА, ИВЕЛИН ИВАНОВ Abstract: The purpose of this article is to follow the change of readability indices in a number of variants of a text developed and transformed by AI. We pay special attention to expressions and phrases that are not suitable for storytelling, that complicate the story and to some useful and interesting ones. We also suggest ways of practising these expressions in foreign language teaching especially by young learners. Keywords: storytelling, AI, readability indices, young learners. Introduction With the development of AI its application in the foreign language classroom is a challenge and a necessity. There are a lot of ways it can aid teachers, especially those who are not much enthusiastic to create their own texts. AI successfully produces materials when given level, topic, style, key words, that supplement the book activities. It also suggests preparing accompanying activities and worksheets. As Sofronieva et al. [6] put it ‘using literary texts to support language learning is a universally incorporated approach’ and their experience with AI-generated texts proved the potential to ‘be successfully used as a pedagogical tool in FLT’. A couple of years ago Schmidt and Strasser [5: 165] declared that ‘data-driven, multi-layered technologies based on algorithms have transformed from a niche discipline into a highly 1 This work is developed as a part of Konstantin Preslavsky University of Shumen project, Gamification in Educational Process, № 08122/06.02.2025. 203 relevant technology for educational, including language learning, purposes.’ And a few years even earlier Kannan and Munday [3: 14] draw the attention to ‘the rising pervasiveness of Artificial Intelligence tools that use Natural Language Processing (NLP) or Automatic Speech Recognition (ASR)’ in terms of the ‘altering how a second language can be taught and learned’. Although some debates about applicability of using AI in foreign language teaching and learning have circulated, literature convincingly promotes a number of possibilities and opens the gate for a plenty of studies on the topic. Our aim is directed to testing some AI tools’ generated versions in terms of their readability and suitability for language teaching. The research on the topic was expanded in some other articles of ours [see 1, 2]. Method Chat GPT 4o free created a story based on the following instructions: Could you please generate a story in English for young learners? Please, include characters from the following traditional stories: Little Red Riding Hood, The Gingerbread Man, The Enormous Turnip. There should be animals and repetition of key phrases in the episodes. Include formulae for greeting people, meeting people, being polite, everyday English. The story The Day Everyone Helped was the result. After the generated text it gave us information about the useful language recycled in it and suggested ‘a printable PDF with illustrations or follow-up activity ideas for classroom use (e.g., acting it out, coloring pages, or a song version).’ We asked it to complicate the story and it developed The Great Forest Picnic. Comparisons about the variants concerning three of the Generative AI tools was an object of a previous work [2]. A further work on the same materials is the aim of the present paper where the focus is on another facet, i.e. readability of the texts obtained. It is considered by means of Gunning Fog index, Coleman Liau index, Flesch Kincaid Grade level, ARI, SMOG, and Flesch Reading Ease. The texts are viewed intelligibility-wise as more comprehensible and sustainable options are suggested to replace where necessary. 204 Each and every of the instruments used for the analysis were free. Readability indices of the AI stories. Looking at readability indices (Table 1) we observe the following data: Table 1. Readability indices in the two Chat GPT stories The Day Everyone Helped The Great Forest Picnic Gunning Fog index 4.12 4.15 Coleman Liau index 0.44 3.24 Flesch Kincaid Grade level 1.75 2.86 ARI -1.11 0.80 SMOG 5.82 6.00 Flesch Reading Ease 91.29 85.20 We see that the difficulty level is higher in the new and more complicated story but with certain indices the difference is minimal (Gunning Fog index and SMOG). This means that if the teacher considers it suitable for a particular group (having in mind language level, interests, need for practise), they can use it successfully. We processed the story The Day Everyone Helped with DeepL and the results concerning the readability changes are shown in (Table 2). The last row presents the index of a similar story created by Grammarly. Looking at Gunning Fog index, Flesch Kincaid Grade level, ARI, SMOG, Flesch Reading Ease, the easiest variant is casual, usually followed by the original. The original is the simplest according to Coleman Liau index. ARI shows all the versions are plain enough, the easiest being casual. According to SMOG they are not very easy. 205 Table 2. Readability indices of the different versions of the story Gunning Fog index Coleman Liau index Flesch Kincaid Grade level ARI SMOG Flesch Reading Ease Original 4.12 0.44 1.75 1.11 5.82 91.29 Simple 4.01 2.16 1.98 0.37 5.90 89.24 Enthusiasti c 4.87 3.06 2.69 0.75 6.51 85.71 Friendly 4.27 2.63 2.13 0.39 6.07 89.82 Casual 3.64 1.95 1.63 0.11 5.57 92.11 Confident 4.14 2.03 2.09 0.31 5.97 88.34 Diplomatic 5.75 4.75 3.69 1.93 7.15 80.67 Grammarly story 5.02 2.79 3.03 1.08 6.56 81.16 If we look at Table 3, we see that the longest story is the diplomatic version, followed by the enthusiastic one and the shortest story is created by Grammarly. All indices point to the diplomatic version as the most difficult (Table 2). As the next difficult story we see Grammarly (with 4 of the indices), the original and the enthusiastic version occur once each. Table 3. Number of characters, words etc. Number of characters Number of words Number of sentences Lexical density Average number of characters per word Average number of syllables per word Average number of words per sentence Original 1,273.00 336.00 68.00 53.57 3.79 1.31 4.94 Simple 1,325.00 321.00 68.00 55.45 4.13 1.33 4.72 Enthusiastic 1,557.00 378.00 68.00 53.17 4.12 1.37 5.56 Friendly 1,521.00 377.00 67.00 53.05 4.03 1.32 5.63 Casual 1,350.00 333.00 68.00 54.95 4.05 1.30 4.90 Confident 1,303.00 316.00 68.00 54.11 4.12 1.34 4.65 Diplomatic 1,955.00 461.00 68.00 47.72 4.24 1.41 6.78 Grammarly story 1,216.00 282.00 64.00 59.57 4.31 1.43 4.41 206 Strange, difficult and useful expressions in the various versions. Sometimes AI gives words and expressions that are not very suitable for a story and we have to edit some parts: e.g. ‘We're trying to get this turnip to work.’ in the simple version instead of ‘We’re trying to pull this turnip.’ in the original. There are expressions that can be edited or simplified: friendly fox and happy cat, we would replace with only fox and cat. Another option is to trust AI and leave the collocations as they were generated bearing in mind that repetition of sounds, regardless of it being a vowel or consonant, has a beneficial effect on holding a chunk in one’s memory [4: 15]. The repeated sounds are easily recognized as [f] in the first case and [æ] in the second one. Moreover, the significance of the latter is undeniable as far as in the Bulgarian students’ mother tongue such a phoneme is missing. It is also worth noting the lack of ordinary logic in the generated text. There is one illogical expression: sunny woods. Usually woods are shady. We can either use woods or shady woods or deep forest. Another thing to mention is that the main character is grandpa. We have either to replace it with an old man, or negotiate Fox’s grandpa or Cat’s grandpa. Other unnecessarily complicated expressions: Enthusiastic version: the ever-determined Little Red Riding Hood, which can be replaced by the Little Red Riding hood ‘It’s a pleasure to make your acquaintance too,’ which can be replaced by ‘Nice to meet you,’ as in the original. They pulled and pulled, and finally, with a great heave, they succeeded in moving the turnip! We skip the marked complex and clumsy expression. AI sometimes uses 1st person singular wrongly: ‘It's so big, it won't come up!’ I'm still stuck, but I'm sure I'll get there soon! And then, just as I was about to give up hope, a tiny mouse ran over! The second line in this excerpt might be the words of the turnip, then we have to add inverted commas as in direct speech. 207 About the third line it is not clear whose words these are: they either belong to the storyteller or to the old man. If it is the Old Man or the turnip, we need inverted commas. The generated story variant introduces an interesting point of view of a turnip who strongly desires to be pulled out. This deviation from the well-known story might turn useful for developing students’ imagination and creativity. As the turnip hopes to escape the earth as soon as possible and perhaps join the company of all those in the long queue who try to help, this can be considered as a pedagogical means of teaching students to help each other especially those in clutch situations. Another possible change in terms of language appropriateness is to replace massive with either big or enormous in the suggested ‘It's absolutely massive! OH LOOK UP!’ The friendly version: the sweet Little Red Riding Hood The whole first sentence is unnecessarily complicated: One lovely sunny morning, the sweet Little Red Riding Hood was taking a walk through the beautiful forest with a basket. The instrument test_and_improve [7] also suggests that this sentence has to be changed. AI uses out of the blue instead of suddenly. The instrument does not like it either. Here again appears the sentence ‘It’s a pleasure to make your acquaintance too,’ They pulled and pulled, but the turnip wouldn't budge. budge is used instead of move. The instrument suggests changes. I'm still not getting anything! instead of Still nothing. Then, the most adorable cat you could imagine came along. We suggest to cross the marked words. This sentence is also marked by the instrument. ‘We're doing our best, but we just can't seem to get that turnip out!’ Here again the instrument suggests changes to be made. We would cross the marked words. 208 Hey, let's grab a snack together! We suggest have instead of grab. This version uses more adjectives and could be used for practicing adjectives with 4th grade students. e.g. lovely sunny morning sweet Little Red Riding Hood beautiful forest we can add a big basket or a basket full of… the most adorable cat, etc. Here are some expressions we would prefer to keep in the story (and practise) despite the fact that the instrument suggests some changes: And they all said: Pull, pull, pull the turnip! You can't catch me, I'm the Gingerbread Man! And then, out came the enormous turnip! Every little helper is important! We're doing really well, thank you so much! They laughed, clapped and cheered. You were so kind and helpful! The first two are an inseparable part of the stories we used. Others can be used in mini dialogues to practise politeness expressions. They all will be acquired during drama and rehearsal activities. The casual version: As stated above, this is the easiest version according to most of the indices. Here we have marked only 4 sentences and our choice is confirmed by the instrument. Out of the blue instead of suddenly but the turnip wouldn't budge: budge instead of move massive instead of big ‘Let's grab a snack together. Fancy a soup?’ Here are some sentences the instrument suggests to change but we would keep and practise as useful expressions: And they all said: Pull, pull, pull the turnip! I'm struggling to get this massive turnip out. 209 You can't catch me, I'm the Gingerbread Man! Every little helper is important! They laughed, clapped and cheered. Everyone fell down – wheeeee! And out came the enormous turnip! You were so kind and helpful! Would you like to walk with me? Run, run, as fast as you can! It's so big, it won't come up! The confident version: ‘Hello! Good morning!’ she declared to the trees and birds. The instrument also suggests this sentence to be changed. The options we suggest: she said to the trees and birds; or she greeted the trees and birds. Here the last sentences are very well formed and could be practised as mini dialogues: ‘Let's have a snack together. Would you like some soup?’ All together: ‘Yes, please! Thank you!’ There is another dialogue from the story that can be used in mini drama activities since it provides a good turn taking practise: 👧 ‘Hello! What's your name?’ Little Red Riding Hood asked. 🏃 ‘I'm the Gingerbread Man! Nice to meet you!’ 👧 ‘Nice to meet you too! Walk with me.’ 🏃 ‘Yes, please!’ Other useful expressions we would keep in the story despite the instrument suggests changes: I can't pull up this enormous turnip. Suddenly, she saw someone running fast! Then, a tiny mouse ran over. We're going to pull this turnip off. A friendly fox then appeared. Let's have a snack together. The diplomatic version: ‘Hello! What is your name?’ Little Red Riding Hood enquired politely. (We suggest ask). 210 ‘I'm the Gingerbread Man! It was a pleasure to make your acquaintance.’ ‘It's a pleasure to make your acquaintance as well! Would you like to walk with me?’ I would be very grateful for any assistance you could provide. If you could try pulling it up, that would be greatly appreciated. They applied a great deal of force At that moment, a friendly fox happened to be passing by. We are endeavouring to navigate this situation. We would be very grateful for any assistance you could provide. is used instead of We’re trying to pull this turnip. Can you help us, please? They expressed their appreciation with laughter, applause and cheers. On the other hand, there are a lot of useful expressions in this version, some of them not suitable for a story (*), but very suitable for mini drama activities: ‘Would you like to walk with me?’ ‘Yes, please!’ a good turn taking practice. Yes, of course! Let's do it together. I'm afraid to say that I have not had any success yet. Then came a happy cat. I would be happy to help! Let's go!’ If you could try pulling it up, that would be greatly appreciated* I'm afraid I'm still experiencing some difficulties* Hello, everyone! Would it be all right if I tried to help too? If you wouldn't mind, could you please look up? And one more interesting and very long and complicated sentence: I would like to think that I am not easily caught, but I am sure that you will be able to catch me if you try.