Sentiment analysis as tool for gender mainstreaming in slum rehabilitation housing management in Mumbai, India

Gender mainstreaming in slum rehabilitation is a critical determinant for the success or failure of it. Slum rehabilitation in Mumbai is a hallmark example of a participatory process which is supposed to improve the quality of life and well-being among the rehabilitated occupants, on paper. Yet our findings show that the key stakeholders’ (i.e. policymaker, architect, management co-operatives and female occupants) sentiments are worlds apart. This gap evokes a need for a more systematic framework for a participatory approach. Verbal communication has the power to implicitly decipher the common emotion or sentiment regarding infrastructure or its related policy. This study uses a quasi-qualitative approach to understand the underlying concerns of gender mainstreaming in slum rehabilitation housing (SRH) management in Mumbai, India. Verbal narratives from semi-structured interviews and focus group discussions are used to explore the concerns of the stakeholders. Sentiment analysis using the machine learning technique of Natural Language Processing (NLP) is used to decode the emotions across the stakeholders. The results indicate that gender inequality in SRH is propagated through lack of communication and the different vocabulary used by the female occupants and the apex policymakers. There are also remarkable differences between male- and female-led co-operatives who are in charge of SRH management. Female-led co-operatives were observed to go beyond basic maintenance duties, to reduce debt and monthly fees and to take up incremental improvements such as decorative elements or installation of solar panels. This participation required a full female body co-operative rather than partial female representation in order to be effective. This study suggests that gender mainstreaming is important not only in SRH design but also in maintenance and that exploring the sentiments of the key stakeholders, as proposed in our analysis, could mitigate undesirable effects of SRH and enable the participatory process.