Article · March 023 doi: 10. 53103/cjlls v3 87 citations reads 951 authors

Another Artificial Intelligence is English Learning Speech Assistant (ELSA) 
English Learning Speech Assistant (ELSA) was created in 2015 by Vu Van and is 
headquartered in San Francisco, California, in the United States. This application uses 
artificial intelligence (AI) and speech recognition to improve English pronunciation. ELSA 
(English Learning Speech Assistant) (English Learning Speech Assistant) Speak is a 
program for learning English that uses speech recognition and artificial intelligence. This 
technology enables a two-way learning process, allowing users to say specific phrases or 
sentences, after which the system analyzes them and offers corrected input. In order to 
identify non-native speakers from most other speech recognition algorithms, the ELSA 
framework is trained using voice samples of persons speaking English with various 
accents. A test is administered to users to measure their degree of proficiency. Most native 
speakers have ELSA ratings of 95 or above, which range from zero to one hundred. ELSA 
can customize the user's learning route by using the findings to identify compelling sounds 
that need further adjustment. Based on the user's current skill level, ELSA will recommend 
courses they should take next. Regarding the teaching of English and the process of 
learning it, Application ELSA (English Learning Speech Assistant) (English Learning 
Speech Assistant) Speak is an application for learning English that uses artificial 
intelligence and voice recognition to assist students in becoming more fluent in spoken 
English. This technology allows for a two-way learning process; for instance, users may 
hear how specific English words are spoken or phrases, after which the algorithm analyzes 
and makes grammatical or structural corrections (Eka, 2020). 
Chatbot 
 
The newest technology created to connect with both people and computers 
automatically are chatbots, one of the artificially intelligent conversational systems (Nghi 
et al., 2019). Previous research has shown the effectiveness of chatbots as a learning tool, 
particularly for English language learners (Afrianto et al., 2019). The created chatbot 
application will function as a conversation buddy for English learners and teachers. In 
general, conversations may be conducted either vocally or in writing, and therefore, the 
user should be able to practice both. Grammar mistake correction and a daily record feature 
are required to support the goal of the chatbot application as a tool for English conversation 
training. The replies received included descriptions of the problems, recommendations for 

---

The Role of Artificial Intelligence 26 
 
words or phrases to replace them, and locations of the faults. This aims to reduce 
grammatical mistakes made by the user. Because of this, the daily log feature helps 
determine how much the user's practice has improved their grasp of English conversation 
and the structure of sentences. Users get more comfortable with conversational English via 
this method and receive feedback on their practice and any errors they may have made. 
The message the user enters is the first item the fundamental chatbot function analyses. 
The chatbot and NLP (Natural Language Processing) could decipher the message's 
meaning by comparing it to the information already included in the database (Haristiani, 
2019).