Suhas Ranganath

3Machine learning papers tracked by Lacuna
Is this you?

Research Overview

Suhas Ranganath focuses on social media analytics, user behavior modeling, and information retrieval. Early work investigated subjective information seeking and political mobilization on platforms like Twitter, drawing on sociological theories to predict user actions. The most recent available research explores multimodal point-of-interest recommendation in location-based social networks.

Research Trajectory

The available evidence outlines a research trajectory active primarily between 2015 and 2017, centered on extracting actionable insights from noisy social media data. Ranganath's early work addressed the challenge of satisfying highly personal and subjective user queries on generic microblogging platforms. In Leveraging Social Foci for Information Seeking in Social Media (2015), Ranganath led the development of a framework to identify which of a user's social connections are best equipped to answer specific information needs. By inferring soft community memberships from Twitter data, this work contributed to the research direction The Challenge of Subjective Ground Truth in Machine Learning, demonstrating how social context can be utilized when objective oracles are unavailable.

Continuing the focus on user behavior modeling, Ranganath investigated political mobilization online. As the lead author of Predicting Online Protest Participation of Social Media Users (2016), Ranganath proposed a predictive framework grounded in sociological theories of protest participation. This research analyzed past status messages and user interactions to forecast whether an individual's subsequent post would declare participation in a protest. This approach aligns with the broader area of Inferring Latent States from Noisy Behavioral Proxies, attempting to push against the limits explored in the theme The Predictability Ceiling in Social and Physical Systems by using historical interaction data to anticipate complex social actions.

In 2017, Ranganath participated as a co-author in research expanding into multimodal data and location-based services. The paper What Your Images Reveal: Exploiting Visual Contents for Point-of-Interest Recommendation examined how photos uploaded to platforms like Instagram reflect user interests and provide descriptive information about specific locations. While earlier work focused primarily on textual and network data, this collaborative effort integrated visual content alongside temporal, geographical, and social signals to improve point-of-interest recommendations.

Current Directions

The most recent paper in the provided evidence is from 2017. Based on this work, Ranganath's latest documented research investigates multimodal recommendation systems within location-based social networks. This direction addresses the integration of visual content—such as user-uploaded images on platforms like Instagram—with established geographical, temporal, and social signals. By extracting latent features from photos, this research aims to better capture both user preferences and the semantic properties of physical locations to enhance point-of-interest recommendations.

Profile updated

See something wrong? Report a mistake