Downloads: 1 | Views: 258 | Weekly Hits: ⮙1 | Monthly Hits: ⮙1
Analysis Study Research Paper | Computer Science & Engineering | India | Volume 12 Issue 4, April 2023 | Rating: 4.6 / 10
Enhancing Query Understanding and Expansion in Retail Services
Yogananda Domlur Seetharama
Abstract: This paper analyzes state-of-the-art systems for generating query suggestions for retail services. This paper also seeks to show that using state-of-the-art algorithms and real-time data processing, query understanding and query expansion approaches can be scaled up to improve the user's search experience. The suggested approach combines all processes into a single computing device, an effective solution for latency problems traditional systems will likely face due to the reliance on third parties. The revolutionary approach uses a structured query index within a search tree to rank query candidates. The temporal scoring function based on the decay using specialties such as sigmoid, exponential, and logarithmic means that the suggestions will be recent and frequently used by the user. This system should also enable multiple retail channels and contain exceptional scores for each channel so that it would suggest a specific product. Analysis by the A/B test suggests that the presented typeahead yields an astonishing 89% improvement over the previous one and has 225ms of latency reduced to 25ms. This paper will conclude by briefly discussing the prospects of the developed system to change the approach to query understanding and query expansion in e-commerce and provide a solid basis for solving inherent problems concerning user satisfaction and sales conversion. Other future work includes expanding on the improvements and incorporating them into other domains to strengthen the proposed system's use.
Keywords: query suggestions, retail services, user search experience, real-time data processing, e-commerce
Edition: Volume 12 Issue 4, April 2023,
Pages: 1926 - 1935