About
TALKO is a chatbot conceptualized and prototyped using Python, and its purpose is not to replace but act as an additional help to the receptionist in the college.
The conversational agent can provide help and
assistance to those looking for information on Faculty, Admissions, the Wellness Center, Equipment to hire, General Information, and more.
assistance to those looking for information on Faculty, Admissions, the Wellness Center, Equipment to hire, General Information, and more.
Concept Video
Summary
For this group project, our first step was to narrow down the area of focus for creating the chatbot. We then sketched and explored various forms the chatbot could take. Once we finalized on projection mapping in addition to a headphone set as the form, I started coding in Python for the chatbot. We then created a low-fidelity prototype to demonstrate our concept. To do this, we scripted an enactment with the prototype and I shot the video for the same.
Setting the Context
The receptionists make sure to solve issues such as managing ID cards, solving students’ queries about different faculty or courses, informing about the locations or changes, etc.
For someone who is new to the building, this can be a little baffling. Moreover students can find it very confusing if there are no receptionists around to answer their queries or inform them about something or the other. Sometimes, some might even prefer to not talk to another person because of their own personal issues and insecurities.
Thus, with the use of a conversational agent, can the reception area be made more user-friendly? Exploring this, the following concept has been finalized upon, after coming up with a few different ideas.
Developing the Concept
TALKO was initially going to be a speaker with a touch screen interface on the top, to enable typed responses, but we figured that since its placed in a public space, the ambient sound might overpower the chat bot or vice versa.
Thus, we came to the conclusion that TALKO could be a voice based assistant in a pair of headphones, and a screen on the reception desk could be utilized using projection mapping, for any text based conversation.
Conceptual Design
Once the Text Button is clicked, a chat can be had with TALKO
To activate it, the user has to speak into the microphone of the headphones that are connected to the desk and are placed at a side for access to the users. Once the words “Hi TALKO” are registered, the start up screen gets projected, allowing the user to use the conversational agent accordingly.
The conversational agent asks the user to pick the category they’d like to talk about including- Faculty, Information, Admissions, Wellness, etc. Given that people aren’t too open about talking to the wellness team or even talking out loud about needing some sort of help, a text based conversation can be had.
The Wellness option facilitates a user to talk to the conversational agent and fix an appointment with any of the counselors available, in a time slot comfortable with the user.
The Faculty option notifies the faculty if a student wants to meet them. It also answers queries about locating some faculty, or even knowing which faculty or administration person to approach for what, etc.
The Information option is basically like an online library of sorts which contains all information about different majors, new notices about other miscellaneous things.
Admissions specifically caters to parents or new people in the college, who are there to inquire about admissions and aren’t students yet. Understanding the process of how college works can be simplified by talking to someone instead of reading a brochure or pamphlet.
Using the Equipment option, one can inquire about availability of equipment like cameras, or speakers, etc and also receive information on the procedure to acquire it.
There is an Other option, for any other queries, where you can talk to TALKO about anything else that these categories seem to not cover, and TALKO shall respond accordingly.
A tick ✔ and a cross ✖ button are always present at the side of the screen. To end a conversation the user has to click the tick icon, after which the user is asked for feedback on whether they are content with the conversation or not. The user can then answer via the tick or the cross again, and TALKO asks if they’d like to have a conversation about anything else. If there has been no conversation from the user’s end for two minutes, TALKO shuts by itself showing a default message.
Finally, we used Python for a low fidelity prototype of the text based conversation with TALKO.
A group project done by Srushti G and Calvin J Stanley