Proposing a new function of a conversational search system that allows natural language communication with users: Searching for Love
Description of new functions
The new function that I propose is called Searching for Love. The primary goal of Searching for Love is to provide users a tailored and interactive search process that goes beyond traditional dating platforms. It utilizes conversational AI to engage users in meaningful conversations, understanding their preferences, values, and compatibility factors, and ultimately facilitate successful matches. Users can also enable Searching for Love to go through the browsing history to find out the hidden web usage patterns that indicate users’ potential preferences for their romantic partners that they might not be aware of.
Searching for Love will work as an extension application for major search engines like Google, Bing, and Baidu. The application will have a conversational user interface. Through engaging conversations with AI, users can effortlessly share their information and preferences, resulting in the creation of a comprehensive dating profile. Beyond interacting with the AI through texts, users are also given the option to activate the camera and microphone to have a “face-to-face” meeting with the AI. Utilizing advanced facial analysis, Searching for Love captures users’ dynamic expressions, generating a series of short video clips that authentically showcase their personality and values — elements that are often challenging to convey through text alone.
By granting Searching for Love access to their browsing history, users unlock personalized suggestions beyond what’s included in their profile. The AI analyzes the user’s web usage and presents relevant information that complements their preferences. To ensure transparency, the AI provides reasons behind each suggestion, derived from the thorough analysis of the user’s online behavior. If users find the suggestions appealing, the AI generates corresponding content to enhance their dating profile. Users retain full control and have the ultimate decision-making power. They can choose whether or not to upload the generated content to their profile, ensuring that their preferences and privacy are always respected.
The illustration of the Search Engine Result Page (SERP) displays user portraits. Each portrait is represented by a bubble, and the size of the bubble corresponds to the compatibility between users — larger bubbles indicate higher compatibility. The clusters of bubbles showcase the reasons behind these matchings, with the most promising matches highlighted in the center of the SERP based on an overall matching score calculation. When a user hovers the cursor over a portrait, the bubble enlarges (the dotted bubble), revealing a snippet of information about that particular user (see Picture 1). When users update their information or preferences, the size and clustering of the bubbles will adjust to reflect these changes.
Grounds for the new functions
1)Incorporating Rich Media in Search Queries:
Ido (2016) highlighted the significance of voice queries, demonstrating that they not only exhibited longer lengths but also featured richer language compared to text-based queries. Building upon this finding, Searching for Love incorporates audio and video elements into the search system, enabling users to express their preferences, values, and compatibility factors in a more nuanced and comprehensive manner. By leveraging rich media interactions, Searching for Love aims to reveal deeper insights into the information needs of users, ultimately facilitating more accurate and tailored matchmaking.
2)Conversational Search Systems for Information Needs Resolution:
Belkin (1980) emphasized that users often face an anomalous state of knowledge when seeking information and may struggle to specify their precise needs. To address this challenge, interactive information retrieval systems, such as conversational search systems, have been developed. Following this paradigm, Searching for Love adopts a conversational search system that goes beyond mere textual interactions. By engaging users in meaningful conversationswith audio and video, this system can guide users towards a better understanding of their own needs. By incorporating natural language processing and machine learning techniques, Searching for Love facilitates personalized and context-aware matchmaking, increasing the chances of finding compatible partners.
3)Optimized Search Result Display:
Granka et al. (2004) observed that users tend to primarily focus on the top few document surrogates when presented with a SERP. To mitigate this limitation and provide users with a broader range of potential matches, Searching for Loveintroduces a novel SERP display format. Instead of traditional textual listings, the SERP showcases clusters of user portraits, with the size of each bubble reflecting the compatibility between users. By enlarging the bubble and displaying a snippet of information when hovering over a portrait, users gain a glimpse into the user’s profile, enabling more informed decision-making during the search for potential matches. This innovative approach aims to maximize search result visibility, improve user engagement, and encourage exploration of a wider range of potential partners.
References
Ido Guy. 2016. Searching by Talking: Analysis of Voice Queries on Mobile Web Search. In Proceedings of the 39th International ACM SIGIR conference on Research and Development in Information Retrieval (SIGIR ’16). Association for Computing Machinery, New York, NY, USA, 35–44. https://doi.org/10.1145/2911451.2911525
Belkin, N. J. (1980). Anomalous states of knowledge as a basis for information retrieval. Canadian journal of information science, 5(1), 133–143.
Laura A. Granka, Thorsten Joachims, and Geri Gay. 2004. Eye-tracking analysis of user behavior in WWW search. In Proceedings of the 27th annual international ACM SIGIR conference on Research and development in information retrieval (SIGIR ‘04). ACM, New York, NY, USA, 478–479. DOI: https://doi.org/10.1145/1008992.1009079
