A webpage for academical experiment that allow building photo collages from images via Google Images API.
Budget: $750 – $1,500 USD
Our team wants to build a webpage for experimental purposes.
We aim to see how people imagine a specific concept without using a text description. Thus, participants in the survey would use images to describe their feelings and understandings of a specific concept. For example, what do they think about a specific Eating experience in a specific country?
In specific, we can describe briefly the project below. Besides, there is a wireframe that we want to have in the end.
Front end requirement:
1. The webpage needs to have a database that stores information about users and their selections on the survey.
2. The webpage database needs to connect with two platforms API: Amazon Mturk (https://www.mturk.com/) and Prolific (https://www.prolific.co/).
The purpose is to make sure once participants complete the survey, and submit their Mturk or Prolific ID, these platforms will reward them accordingly.
UX flow: Please check the demo flow here along with the below description: https://dotua6.dreamhosters.com/exporthtml/index.html
3. Initially, when users go into the webpage for experiments, they would choose to what extent they are familiar with the concept. For example, Dining experience in Norway, Sweden, .etc.
4. After the participant's rating, we will pick which concept they are most familiar with to direct them to the relevant concept.
5. For example, if User A chooses Norwegian Dining Experience. He will be asked to use a search box to find relevant images that describe his concept or his imagination about the Norwegian Dining Experience. Relevant images here would be pulled from Google Cloud Images. The images provided by search would be limited to excluding graphics that directly point to a specific country. For example, for the Norwegian Food experience, we don't want that people can search for a Norwegian flag or something can point directly to Norway. The purpose is to let them express their perception of food without thinking about Norway alone.
Each image will contain tags and alt text, or image default description (available from Google).
6. A user will find several relevant pictures, drags them, and drops them into a white space to create a photo collage.
7. Finally, after completing the photo collage, the user will write a short description of the collage and will submit it. The submitted collage along with image tags and default descriptions will be sent to the Clarifai platform for analysis (https://clarifai.com/) via the API of Clarifai.
Data collection:
8. Each participant joining the survey would have each data point recorded for analysis purposes, including which choices, options they chose or images with tags, and photo collages they built. Please kindly check the detail below:
Data collected requirements:
0. Survey answers
1. The time participants spent on each task
2. How many search attempts, keyword searching to find images
3. Frequent of the taggings in images, Top and least words, compose the short sentences based on the taggings.
4. Graphic dimensions for each image in the collage
Extra: can we limit the display image storage (e.g., show only Norwegian food)? Will the website show the updated result immediately (on the go) or compare btw conditions?
Please kindly check the Detail of the data analysis in the attachment.
Best regards,
We aim to see how people imagine a specific concept without using a text description. Thus, participants in the survey would use images to describe their feelings and understandings of a specific concept. For example, what do they think about a specific Eating experience in a specific country?
In specific, we can describe briefly the project below. Besides, there is a wireframe that we want to have in the end.
Front end requirement:
1. The webpage needs to have a database that stores information about users and their selections on the survey.
2. The webpage database needs to connect with two platforms API: Amazon Mturk (https://www.mturk.com/) and Prolific (https://www.prolific.co/).
The purpose is to make sure once participants complete the survey, and submit their Mturk or Prolific ID, these platforms will reward them accordingly.
UX flow: Please check the demo flow here along with the below description: https://dotua6.dreamhosters.com/exporthtml/index.html
3. Initially, when users go into the webpage for experiments, they would choose to what extent they are familiar with the concept. For example, Dining experience in Norway, Sweden, .etc.
4. After the participant's rating, we will pick which concept they are most familiar with to direct them to the relevant concept.
5. For example, if User A chooses Norwegian Dining Experience. He will be asked to use a search box to find relevant images that describe his concept or his imagination about the Norwegian Dining Experience. Relevant images here would be pulled from Google Cloud Images. The images provided by search would be limited to excluding graphics that directly point to a specific country. For example, for the Norwegian Food experience, we don't want that people can search for a Norwegian flag or something can point directly to Norway. The purpose is to let them express their perception of food without thinking about Norway alone.
Each image will contain tags and alt text, or image default description (available from Google).
6. A user will find several relevant pictures, drags them, and drops them into a white space to create a photo collage.
7. Finally, after completing the photo collage, the user will write a short description of the collage and will submit it. The submitted collage along with image tags and default descriptions will be sent to the Clarifai platform for analysis (https://clarifai.com/) via the API of Clarifai.
Data collection:
8. Each participant joining the survey would have each data point recorded for analysis purposes, including which choices, options they chose or images with tags, and photo collages they built. Please kindly check the detail below:
Data collected requirements:
0. Survey answers
1. The time participants spent on each task
2. How many search attempts, keyword searching to find images
3. Frequent of the taggings in images, Top and least words, compose the short sentences based on the taggings.
4. Graphic dimensions for each image in the collage
Extra: can we limit the display image storage (e.g., show only Norwegian food)? Will the website show the updated result immediately (on the go) or compare btw conditions?
Please kindly check the Detail of the data analysis in the attachment.
Best regards,