Life on the 4th Floor

For people who’ve known me all through my stint at IIIT-Delhi, the title of this blog should come as no surprise. I was infamous for spending a good portion (read: majority) of my time sitting in/around the 4th floor of the (old) Academic building to the extent that I have been called ‘4th floor boy’ at multiple points. The 4th floor of the Old Academic Building was home to the Precog lab (at the time) and was the primary reason for me spending most of my waking hours in this part of campus. While this may seem excessive (also slightly exaggerated), I can confidently say that my time at Precog has been full of the most defining experiences I have had in my life.

My journey at Precog started pretty early. Towards the end of my first semester, we got an email from PK talking about his group and that he’s looking for second/third year undergrads to join Precog. Being extremely under qualified, I decided to shoot him an email asking if I could somehow help out. To my surprise he seemed interested and and asked me to send my resume. Brimming with naïveté, I responded to him saying “I don’t have a resume but here’s 4 paragraphs of what I’ve done instead”. I am genuinely still surprised (and grateful) that PK didn’t delete this email thread right there, but instead decided to connect me with Nilaksh (Precog alum, now doing a PhD at Georgia Tech) for an eligibility “task”. In hindsight, I’m very glad that the naive 18 year old in me had the drive to reach out to PK that early on because sending that one email impacted the course of my life in a monumental way. I know it feels a little dramatic, but trust me when I tell you that getting involved with Precog was probably one of the best decisions I have ever made.

In the beginning, it was extremely intimidating since I was suddenly surrounded by very smart and experienced people. We had weekly meetings where we would each talk about what we worked on in the week and would give each other constructive feedback. The first couple of times, I walked out these discussions with a much lower self-esteem than when I walked into it; it was overwhelming to be interacting with people who have such a deep understanding of the domain they’re working in and it was easy to feel clueless. A few weeks in, once I had the chance to settle in, things felt a lot different. Things started to seem a little more familiar and I felt as if I was actually starting to contribute to the discussions. The “intimidating” PhD students actually turned out to be the nicest and most helpful people I have had the opportunity of working with, complete opposite to my first impressions. I would (very liberally) bother them with questions about my project and they would actually invest a lot of time trying to help me through. I cannot stress on how important our interactions were, and how big a role they played in my time there (more on this later).

During my internship at MPI, summer 2017.

At the risk of sounding immodest, I would like to think that I did some pretty cool things during my undergrad. I interned at Georgia Tech (Atlanta, US) in the summer of my sophomore year and then at Max Planck Institute for Software Systems (Saarbruecken, Germany) the next summer, which were incredible opportunities from both, an academic, as well as life experience standpoint. I was also lucky enough to author some very interesting publications, including a first-authored publication at CHI which is a top-tier conference. The point behind my bragging is that all of this (and a whole lot of other things I haven’t mentioned) were possible single-handedly because I was working at Precog. I got a lot of support from both PK and the other lab members which made it possible for me to visit these places and have a high work output there. In my opinion, I think Precog is one of the very few research groups in the country where the quality of work/connections make it possible to get these kind of opportunities; I am very confident that I wouldn’t have as much to (humble) brag about if I wasn’t a part of this group.

I went on to work with Precog for over 3.5 years which is much longer than the usual amount of time people work with research groups at IIIT-Delhi. One of the biggest selling points (and why I continued working at Precog) is the real world impact Precog has as a group. Unlike a lot of research groups where the job is “done” once a paper is published, I have personally seen many different tools/technologies which have come from said research, and vice versa (papers coming from tools). Being an impact driven person, this was very motivating for me and was a big reason for me working with Precog for as long as I did. Even when I joined as a first semester undergrad, I was put on a team working on a mission-critical project of national importance. If you are someone who’s interested in using computer science to impact human lives on a large scale, I urge you to take a moment and check Precog out.

When students and advisor decide to wear the same GaTech Tshirts!

It was only much later, when I was fortunate enough to visit some of the best CS schools across the globe, that I got a chance to draw a parallel between how things are at Precog vs the rest of the world. I had expected to be completely blown away by the resources/infrastructure/functioning of these top research groups, but honestly a lot of it felt very familiar. It was only then that it slowly started to realize how representative Precog is of a leading research group even though we’re based in a small school in India. In all my time there, there has never been a moment where a project was “slowed down” because of a lack of computational resources; it was actually quite the opposite, where a project sometimes (mostly) had an entire server with multiple GPUs to themselves. Put together the fact that almost every project has collaborations with industry/other top research labs across the country (or globally), and you don’t feel like you’re in a small institute in India anymore.

While the resources, infrastructure, connections etc. are great, the people at Precog are the real “wealth”; Precog is less of a “research group” and more of a family. I know how cheesy that sounds, but honestly there is no other way to put it. I have personally seen a lot of members come and go from the group, and the one thing every single person will tell you is how “together” everyone is. Everyone knows what every other person is working on, and there is a very high likelihood that most people would have contributed to your project in one way or the other. This culture of helping each other is unique; I have never seen any other group that is so closely linked to each other and that spends so much time working with other people to augment their work. I personally owe a lot to so many members of the group who I bothered continuously (and still continue to bother, I literally just messaged some of them a couple of minutes ago) and who never said no, even though they probably had a lot of other things going on. (Big shoutout to (Dr) Prateek, (Dr) Niharika, Anupama, Srishti, Megha, Rohan, Archit, Nilaksh, Yatharth and every one else I had the opportunity to work with, safe to say I wouldn’t be half as successful at what I do if it wasn’t for all of you guys).

Partying!

I realize that I’ve only talked about the work side of the culture at Precog, but there is also an equally important side to the culture where we like believe in a life beyond work. From “official” socials and organised trips, to (very) impromptu plans and celebrations, we have done it all. Some things that stand out for me include memorable Precog birthday celebrations, as well as celebrating individual birthdays or paper acceptances in an elaborate way. Over the years, I’ve spent a significant amount of time at Precog just hanging out and chilling with people; I distinctly remember there was a time where we would go out partying together at least twice a week.

One of those “socials”!

I’ve talked a lot about the group as a whole, but what you need to understand is that none of this would have been possible without PK. Without a doubt, he is the one person who makes this group function the way it does and I don’t think there’s anyone who could do it better. PK has been one of the biggest supporters I’ve had all through my time at IIIT-Delhi (and now even after), even much before he officially became my advisor. He has been one person that has gone above and beyond to help me push myself (or dig myself out of a hole), and trust me when I say that this happened a lot. Even though he has *a lot* of things going on, he somehow always makes time for his students and makes sure that things are going well at their end. I could go into a lot more depth on how big of a (positive) influence PK has had on my life, but I will stop at saying that I am truly amazed by how a person can alone make such a big impact on so many people. To the say that I am glad I had the opportunity to work with him and interact with him is a major understatement.

First of all, I need to apologize for this entire blog so far. I read through this a couple of times but I’m still very unhappy with how it has come out. Part of it has to do with my literary ineptness, but it’s mostly because it is (personally) impossible to write a blog post describing how awesome this group and everyone in it are. The downside of spending so much time working with a group of people is that you have too many experiences and memories for you to be able to pen down in a couple hundred words. I could probably go on for hundreds of pages and still not do justice to my experience here. Precog is a place where I had so many essential learning experiences (in all meanings of the term) and a lot of fun. There have been ups and downs (which is true for everywhere) but I cannot think of a single moment over the years where I regretted being a part of the group. It’s been a couple of months now since I left Precog (and IIIT-Delhi) to join the MS CS program at Carnegie Mellon University, but I can confidently say that my time here has played a major role in enabling me to be successful in life on the outside.

Given a choice now, would I do it all over again? In a heartbeat.

What’s your MWI? : A social media based study of Mental Well-Being in college campuses

College students’ mental health concerns are a persistent issue; psychological distress in the form of depression, anxiety and other mental health challenges among college students is a growing health concern. However, very few university students actually seek help related to mental illness. This arises due to various barriers like limited knowledge about available psychiatric services and social stigma. Further, there is dearth of accurate, continuous and multi-campus data on mental well-being which presents significant challenges to intervention and mitigation strategies in college campuses.

Recent advances in HCI and social computing show that content shared on social media can enable accurate inference, tracking and understanding of the mental health concerns of users. There has also been work showing that college students appropriate social media for self-disclosure, support seeking and social connectedness. These facts, coupled with the pervasiveness of social media among college students, motivated us to examine the potential of social media as a “measure” for quantifying the mental well-being in a college population. Specifically, we focused on the following research goals:

  • Building and validating a machine learning model to identify mental health expressions of students in online communities
  • Analysing the lingusitic and temporal characteristics of the identified mental health content
  • Developing an index for the collective mental well-being in a campus, and examining it’s relationship with university attributes like academic prestige, enrollment size and student body demographics

We obtained a list of 150 ranked major universities in the US by crawling the US News website. We also obtained university metadata like gender distribution, tuition/fee during this crawl. Next, we crawled the Wikipedia pages for these 150 universities for extracting the student enrollment, type of university (public/private) and the setting (city/urban/suburban/rural) at each institute. Lastly, we obtained information on the racial diversity at each university from an article on Priceonomics. We study these universities in our work and use the metadata in our analysis.

Geographical distribution of the 150 ranked universities we studied

For social media data, we focus on Reddit. Reddit is known to be a widely used online forum and social media sites among the college student demographic. It’s forum structure allows creation of public online communities (known as “subreddits”), including many dedicated to specific college campuses. This allowed us to collect a large sample of posts shared by university students in one place. Although Facebook is likely more popular/widespread among students, it is challenging to use Facebook in such studies since the content shared is largely private, making it challenging to obtain such large data from it. Further, the semi-anonymous nature of Reddit enables candid self-disclosure around stigmatized topics like mental health.

After a manual search for subreddits for each university, we were able to identify public subreddit pages for 146 of the 150 universities. Next, we focused on correcting the “under-adoption” bias in subreddits. Subreddits which had a small fraction of Reddit users (as compared to university enrollment) were filtered out due to being under-representated. This left us with 109 universities with adequate Reddit representation. We leveraged the data on Google BigQuery (combined with some additional data collection) to get all posts ranging from June 2011 to February 2016. The final dataset used for our analysis included 446,897 posts from 152,834 unique users.

Since Reddit data does not contain any gold standard information on whether a post in a university subreddit is a mental health expression, our first goal was to use an inductive transfer learning approach to build a model to identify such content in a university subreddit. First, we include (as ground truth data) Reddit posts made on various mental health support communities. Prior work has established that, in these communities, individuals self-disclose a variety of mental health challenges explicitly. We use these posts as the “positive” posts and, parallelly, we utilize another set of Reddit posts, made on generic subreddits unrelated to mental health, as “negative” posts. We obtain 21,734 posts for each category, which we use as the positive and negative class for building a classifier. We observed a validation accuracy of 93% and an accuracy of 97% on a test set of 500 unseen, expert-annotated posts from our university subreddit data. We then proceeded to use this classifier for labelling the 446,397 other posts across the 109 university subreddits. Our classifier identified 13,914 posts (3.1%) to be mental health expressions, whereas the rest of the 432,483 posts were marked not about the topic. This corresponded to 9010 unique users out of a total of 152,834.

Next, we looked at the linguistic characteristics of the posts identified to be mental health expressions by conducting a qualitative examination of the top n-grams uniquely occuring in these posts. We found that students appropriate the Reddit communities to converse on a number of college, academic, relationship, and personal life challenges that relate to their mental well-being (“go into debt”, “doing poorly in”, “only one homework”, “up late”, “the jobs i”). The n-grams also indicated that certain posts contained explicit mentions of mental health challenges (“psychiatric”, “depression”, “killing myself”, “suicidal thoughts”), as well as the difficulties students face in their lives due to these experiences (“life isnt”, “issues with depression”, “was doing great”, “ruin”, “cheated”). Some of the top n-grams were also used in the context of seeking support (“need help”, “i really need”, “could help me”).

For the temporal analysis of mental health content, we first study the proportion of posts with mental health expression across the years. The figure below shows the content per year (along with a least squares line fit). We observed that the proportion of posts with mental health expressions has been on the rise — there is a 16% increase in 2015, compared to that in 2011.

We then looked at how this trend varies over the course of an academic year. The plots below show the trend separately for universities following the semester system and the quarter system. Between August and April, for the universities in the semester system, we observed an 18.5% increase in mental health expression; this percentage was much higher: 78% for those in the quarter system, when compared between September and May. On the other hand, we observed a reverse trend in mental health content during summer months, for both semester and quarter system universities.

Trend in the semester months
Trend in the quarter months
Trend in the summer months for quarter system universities

 

 

Trend in the summer months for semester system universities

 

 

 

Lastly, as a part of our third research goal, we formulated an index we refer to as the Mental Well-Being Index (MWI), as a measure of the collective mental well-being in a university subreddit, based on the posts labelled as mental health related by the classifier. We then computed the MWI metric for all 109 subreddits and examined it’s relationship with the university attributes.

 

 

 

 

 

By visualising these relationships (as above), we gleaned several interesting observations. We found:

  •  Universities with larger student bodies (enrollment) as well as greater proportion of undergraduates in their student bodies tend to be associated with lower MWI
  • MWI of the 66 public universities we consider, is lower, relative to that in the 43 private universities, by 332%
  •  MWI is lower in the 7 rural and 33 suburban universities by 40-266% compared to others, while it is the highest in the 31 universities categorized to be in cities (by 29-77%)
  • Universities with higher academic prestige (or low absolute value rank) and higher tuition tend to be associated with higher MWI
  • MWI tends to be lower in universities with more females (or sex ratio, male to female <= 1) by 850%

Further, although our data shows a marginally lower MWI in universities with greater racial diversity, we did not find statistical significance to support this claim.

Our work here (the complete paper accepted at CHI 2017) further details our analysis in depth. Below is an infographic for our work.