Checkpoint 1
Introduction: Instagram might be the most popular social media application in our generation. A common question asked before posting a picture is what the caption should be. There are a lot of different types of captions including descriptive ones, inside jokes, puns, quotes, and short phrases. Our goal is to create a model that given a picture can output a suitable instagram caption. In order to do this we will create a classification that is trained on real instagram pictures and corresponding captions to learn what sort of captions are most appropriate for which pictures.
Challenges: What has been the hardest part of the project you’ve encountered so far?
So far we have worked on just getting our data off instagram and are beginning to parse it. We looked into a few APIs to scrap images and captions from instagram. We want to get data from the top 100 English language account users that are verified but it has been challenging to ensure we are scraping from the specific accounts we want and getting a sufficient amount of data from each account so that after cleaning the data we still have enough information. We believe that parsing the captions will be the most difficult part as we have to tokenize on letters. Also cleaning the data is challenging as it requires removing pictures without captions and posts that have more than one picture.
Insights: Are there any concrete results you can show at this point? How is your model performing compared with expectations?
We don’t have any concrete results of running the model and getting some output. Since all we have really done is gathered data and tried to figure out the best way to parse it, the only results we have are just getting instagram data and seeing that we can access it correctly.
Plan: Are you on track with your project? What do you need to dedicate more time to? What are you thinking of changing, if anything?
Yes, we believe we are on track with the project. By far the most difficult part of the project should be the gathering and parsing of the data, so once we are sure that is done, we think we will be in very good shape to finish. We have planned out the model architecture which combines our work from past assignments. After completing that, all we would then have to do to reach our base goal is likely print out some results. We then plan to play with the architecture a little and tweak hyperparameters to get to the target goal and then maybe the stretch goal, if we think we can.
Log in or sign up for Devpost to join the conversation.