SNEHAI

New

An innovative chatbot in India that was purposefully conceptualized and designed by the Population Foundation of India to educate and inspire adolescents and young adults to live healthy lives, promote sexual and reproductive health (SRH), and advocate for health and well-being of women and girls.

Read On
Contact

The Population Foundation of India

CASE STUDY SOURCE

The Population Foundation of India

PROJECT STATUS

ONGOING

IMPLEMENTATION PARTNERS

The Population Foundation of India

CONTACT (AS LISTED IN ORIGINAL CASE STUDY)

Hua Wang

Enabling Environment Building Blocks

Services and Applications

Family Planning Program Classification

Demand Generation

An innovative chatbot in India that was purposefully conceptualized and designed by the Population Foundation of India to educate and inspire adolescents and young adults to live healthy lives, promote sexual and reproductive health (SRH), and advocate for health and well-being of women and girls.

Introduction

With a population of approximately 1.4 billion, India accounts for about 18% of all people on the planet [1] with half of this population being under the age of 25 years. [2] Despite policy commitments and significant strides made in recent years, the informational needs of adolescents and youth are poorly met, quality education about SRH is highly limited, contraceptive practices are heavily skewed toward female sterilization, and unsafe abortions are rampant. [3-6] In particular, young people in India have limited awareness of contraception and sexually transmitted infections; their knowledge base consists of inaccurate information, and their family life education is highly insufficient. [7]

Although the government of India endorsed a national adolescent health program to support access to SRH services since 2014, direct contact with frontline health workers, even by married young women, remained extremely low. In addition, contact with unmarried youth and the use of SRH services at adolescent-friendly health clinics are almost completely amiss. [8] Uncomfortable and embarrassed to ask, young people in India have increasingly referred to web-based platforms to look for answers to SRH questions and have garnered misleading or incorrect information. [9,10]

At the beginning of 2020, India boasted of 1.1 billion mobile phone connections, covering 78% of the population. [11] With attractive pricing from India-based telecom giants, such as Jio, internet penetration and social media use through mobile networks are rapidly growing. [11,12] Facebook (Meta platforms) is an obvious leader in the social media space in India, with 320 million users. [13] With this massive expansion of information infrastructure comprising wireless networks, digital technologies, and social media, Indian youths, both in urban and rural areas, are increasingly being plugged into this technology web. The SnehAI chatbot aims to provide a safe space for Indian youth to have conversations about SRH, dispel sex-related myths and taboos, offer accurate information about safe sex and contraceptive choices, and address mental health concerns.

About SNEHAI

Chatbots are automated nonhuman agents that engage in conversations with human actors [14]. By design, the user experience in a chatbot strives to be pleasant, as it mimics a scenario in which 2 humans are talking with each other. A chatbot responds by accessing information stored in large digital data repositories. Chatbots quickly sieve what is relevant and convert programming codes into expressions that humans can understand. Although chatbots are often text-based, their capabilities have exploded since the pioneering program ELIZA, an early natural language processing computer program [15], especially in their increased sophistication and accuracy in understanding natural language using artificial intelligence (AI) technologies [16,17]. Machine learning, a method of data analysis that automates analytical model building, makes the deployment of AI immensely valuable for disseminating information.

SnehAI represents the first Hinglish (Hindi + English).

AI chatbot that is co-created with and for young people, especially those from vulnerable sections of society, to deliberately facilitate communication about SRH topics and promote social and behavioral change [18,19]. SnehAI provides an unusual repository of educational knowledge in an entertaining and engaging container, commonly referred to as entertainment-education or edutainment [20,21] In April 2020, SnehAI (version 2.0) was launched with a natural language processing (NLP) platform and better content flow to make it more intelligent in conversations with users [19]. It is required that an individual be a user of Facebook Messenger to chat with SnehAI

EVALUATION AND RESULTS

The aggregate dashboard analytics data on SnehAI user behaviors was examined over a period of 5-month. A total of 8,170,879 messages were exchanged between SnehAI and 135,263 unique chatbot users, including 5,100,449 (62.42%) outgoing messages from the chatbot to the users and 3,070,430 (37.58%) incoming messages from the users to the chatbot. The ratios between outgoing and incoming messages were consistently around 60:40 over time, with the exception of repeated outgoing messages in July (such as technical errors). Specifically, incoming messages from the users to the chatbot were sent through free texts (1,599,339/2,878,908, 52.09%); clickable reactions (1,279,569/2,878,908, 41.67%); or GIFs, audios, images, and videos (191,522/2,878,908, 6.24%).

The average user engagement with SnehAI was 1.9 sessions, 7.6 minutes, and 56.2 messages exchanged. The time spent by top users increased from 2 to 3 hours in the first 3 months to 7 hours in 14 sessions in August and 14 hours in 47 sessions in September.

User engagement was also tracked through the content categories, including both guided flows using clickable reactions and the handling of free-text messages using NLP. From onboarding to learning about SnehAI, to its privacy policy, and its main menu to the videos, stories, games, helplines, and query responses through NLP, the chatbot users traversed across these content areas for 1,430,416 times over the course of 5 months. Approximately half (705,305/1,430,416, 49.31%) of these interactions were about the chatbot responding to the user queries, including small talks and any questions related to themes of health communication. The next highest frequency content category was user enrolment (257,042/1,430,416, 17.97%), about SnehAI (115,131/1,430,416, 8.05%), privacy policy (96,305/1,430,416, 6.73%), main menu (83,692/1,430,416, 5.85%), helplines (71,211/1,430,416, 4.98%), stories (61,582/1,430,416, 4.31%), games (20,897/1,430,416, 1.46%), and videos (19,251/1,430,416, 1.35%). These trends in content engagement distribution were relatively consistent (with occasional fluctuations).

Our analytics tracking data showed a count of 99,936 typed text messages from the users that SnehAI handled and responded through the NLP system with queries related to the six topical themes

safe sex practices, such as consent, frequency of sexual intercourse, oral and anal sex, impact of other health ailments on sex life, and unplanned pregnancy (57,158/99,936, 57.19%); choice of family planning methods, such as male and female condoms, oral contraceptive pills, intrauterine devices, injectables, and SRH-related themes, such as abortion, sexual intercourse during and after pregnancy, polycystic ovarian disease, and infertility issues (6287/99,936, 6.29%); female reproductive health concerning menstruation (e.g., regularity, pain, discharge, and spotting), virginity, and premarital sex (13,965/99,936, 13.97%); adolescent sexual health issues, such as nightfall, masturbation, pornography, sexual stamina, erectile dysfunction, and STIs (15,160/99,936, 15.17%); adolescent mental health issues regarding peer pressure and bullying (2343/99,936, 2.34%); and nutrition and social determinants of health, such as child marriage and gender equality (4623/99,936, 4.63%).

SnehAI chatbot user gender-disaggregated data revealed an extreme gender gap

among the unique chatbot users over the 5-month period, 93% (125,795/135,263) were male, 6.8% (9198/135,263) were female, and 0.2% (270/135,263) unknown. This gender ratio was disproportionally skewed toward male users compared with that in the United Nations Development Programme report on the gender distribution of internet users (29% female) and Facebook users (22% female) in India [12]. This might be a function of female users facing gender disparities in mobile device ownership and low digital media literacy in India. Also, the 15,000 queries that were sampled by the data science team established a behavioral pattern that revealed that female users exhibited significantly lower self-confidence in their conversations with SnehAI

Lessons Learned

  • SnehAI pioneered the use of an AI chatbot to engage users, especially young men, in conversations about SRH. A total of 8.2 million messages were exchanged in a 5-month period, with almost half of the incoming user messages being free-text queries comprising personal questions and concerns about SRH.
  • The SnehAI avatar based on Dr. Sneha—the protagonist in Main Kuch Bhi Kar Sakti Hoon television show—presented itself as a trustworthy friend and mentor.
  • Furthermore, the curated content from this gender equality and family planning initiative was found to be both entertaining and educational. Also, the NLP system worked smoothly and effectively to personalize the chatbot response and optimize user experience.
  • The results of our instrumental case study provide evidence that SnehAI offers a safe space for users to talk about sensitive SRH topics, seek and obtain accurate information, access services locally through helplines, and seek personal counseling.
  • The patriarchal restrictions on women’s autonomy, mobility, and self-expression in India were vividly reflected in SnehAI user analytics data. This disparity in the ratio is a risk and can introduce gender bias in the AI chatbot’s NLP model.
  • The team is currently adapting field-based promotional strategies to include more female users.

Conclusion

As the first Hinglish (Hindi and English) AI chatbot deliberately designed for social and behavior change communication, SnehAI is an innovative, unique, and promising app for engaging vulnerable and hard-to-reach populations groups in the context of SRH education and discussion. It offered a private, non-judgmental, and safe space for users to talk about otherwise sensitive topics, obtain accurate and trustworthy information, access national services through toll-free helplines, and seek personalized consultations. It also opened up exciting opportunities to leverage the emotional appeal of storytelling through thoughtful yet entertaining content, positive outlook of an avatar, relatable verbal and nonverbal expressions, and friendly tone of voice to effectively engage young people in talking and learning about SRH. SnehAI will further be scaled up to reach potential beneficiaries within and beyond India.

References

  1. India population (2021) Worldometer. [2021-04-22]. Available online: https://bit.ly/3oWivYn [Last accessed 17th February 2022]
  2. Sample Registration System Statistical Report 2018. Office of the Registrar General & Census Commissioner, India. [2021-04-22]. https://bit.ly/3HVXowX [Last accessed 17th February 2022]
  3. Jejeebhoy S, Santhya K, Zavier A.(2020) Sexual and reproductive health in India. Global Public Health. 2020. [2021-04-22]. Available online: https://bit.ly/3GZUUw6 [Last accessed 17th February 2022]
  4. Muttreja P, Singh S. (2018) Family planning in India: the way forward. Indian J Med Res. 2018 Dec;148(Suppl):1–9. doi: 10.4103/ijmr.IJMR_2067_17. Available online: https://bit.ly/3sJr6ip [Last accessed 17th February 2022]
  5. Tripathi N. Does family life education influence attitudes towards sexual and reproductive health matters among unmarried young women in India? PLoS One. 2021 Jan 25;16(1):e0245883. Available online: https://bit.ly/3BqkfhM [Last accessed 17th February 2022]
  6. National Family Health Survey (NFHS-4), 2015-2016. Dhsprogram. [2021-04-22]. Available online: https://bit.ly/3oUtvpj [Last accessed 17th February 2022]
  7. UDAYA (2016) Understanding the lives of adolescents and young adults: Bihar factsheet 2016. Popcouncil. [2021-04-22]. Available online: https://bit.ly/3BGrVwB [Last accessed 17th February 2022]
  8. Adolescent health: priorities and opportunities for Rashtriya Kishor Swasthya Karyakram (RKSK) in Uttar Pradesh. Population Council, New Delhi. (20220 Available online: https://bit.ly/3oU7Qxo [Last accessed 17th February 2022]
  9. Gadkari RP, Somani G, Nayak CS, Giri AS. (2012) A study for sexual health awareness in the adolescent population (13-18 years) attending dermatology OPD. Indian J Sex Transm Dis AIDS. 2012 Jul;33(2):148–9. Available online: https://bit.ly/3JylpdN
  10. Muzammil K. (2019) Helplines for adolescents in India. In: Mahmood SE, editor. Concise Text-Book of Adolescent Health. New Delhi, India: IP Innovative Publication Pvt. Ltd; 2019. pp. 108–13.
  11. Kemp S. Digital 2020: India. Datareportal. 2020. [2021-04-22]. Available online: https://bit.ly/3JtMMWg
  12. UNDP (2021) Social media for youth and civic engagement in India. Available online: https://bit.ly/3oTI8ZO [Last accessed 17th February 2022]
  13. Statista (2021) Social media usage in India-statistics and facts. Available online: https://bit.ly/3JATSIE Last accessed 17th February 2022]
  14. Følstad A, Araujo T, Papadopoulos S, (2019) Chatbot research and design. Proceedings of the Third International Workshop, CONVERSATIONS 2019; Third International Workshop, CONVERSATIONS 2019; November 19–20, 2019; Amsterdam, The Netherlands. 2019
  15. Weizenbaum J. ELIZA-A computer program for the study of natural language communication between man and machine. Commun ACM. 1966 Jan;9(1):36–45. doi: 10.1145/365153.365168.
  16. Seymour M, Riemer K, Kay J. (2018) Actors, avatars and agents: potentials and implications of natural face technology for the creation of realistic visual presence. J Assoc Inf Syst. 2018:953–81. doi: 10.17705/1jais.00515.
  17. Stoeckli E, Dremel C, Uebernickel F (2019) Brenner W. How affordances of chatbots cross the chasm between social and traditional enterprise systems. Electron Mark. 2019 Aug 08;30(2):369–403. Available online: https://bit.ly/3HZtFDo [Last accessed 17th February 2022]
  18. Population Foundation of India Outreach and communication engagement using AI-powered chatbot. (2019). A Development and Pre-Testing Report Submitted to the United Nations Population Fund; 2019; India. 2019.
  19. Population Foundation of India SnehAI: an AI-powered companion for young people. A Program Report Submitted to Centre for Social and Behaviour Change; 2020; India. 2020
  20. Singhal A, Rogers E. (2012) Entertainment-Education: A Communication Strategy for Social Change. Mahwah, New Jersey: Routledge.
  21. Singhal A, Cody M, Rogers E. (2004) Entertainment-Education and Social Change: History, Research, and Practice. Mahwah, New Jersey: Erlbaum Associates.

Other Resources

Wang H, Gupta S, Singhal A, et al. (20220 An Artificial Intelligence Chatbot for Young People’s Sexual and Reproductive Health in India (SnehAI): Instrumental Case Study. J Med Internet Res. 2022;24(1): e29969. Published 2022 Jan 3. Available online: https://bit.ly/3sICCub [Last accessed 19th March, 2022]