{"id":934,"date":"2023-12-22T17:34:56","date_gmt":"2023-12-22T17:34:56","guid":{"rendered":"https:\/\/www.wsdm-conference.org\/2024\/?page_id=934"},"modified":"2024-03-05T21:27:25","modified_gmt":"2024-03-05T21:27:25","slug":"industry-day","status":"publish","type":"page","link":"https:\/\/www.wsdm-conference.org\/2024\/industry-day\/","title":{"rendered":"Industry-day"},"content":{"rendered":"\n

Keynote speakers<\/mark><\/h2>\n\n\n\n

Delphic Costs and Benefits in Web Search: A utilitarian and historical analysis<\/mark><\/h3>\n\n\n\n
<\/div><\/div><\/div>\n\n\n\n

Abstract:<\/strong>
We present a new framework to conceptualize and operationalize the total user experience of search, by studying the entirety of a search journey from an utilitarian point of view.

Web search engines are widely perceived as \u201cfree\u201d. But search requires time and effort: in reality there are many intermingled non-monetary costs (e.g. time costs, cognitive costs, interactivity costs) and the benefits may be marred by various impairments, such as misunderstanding and misinformation. This characterization of costs and benefits appears to be inherent to the human search for information within the pursuit of some larger task: most of the costs and impairments can be identified in interactions with any web search engine, interactions with public libraries, and even in interactions with ancient oracles. To emphasize this innate connection, we call these costs and benefits Delphic, in contrast to explicitly financial costs and benefits.

Our main thesis is that users\u2019 satisfaction with a search engine mostly depends on their experience of Delphic cost and benefits, in other words on their utility. The consumer utility is correlatedwith classic measures of search engine quality, such as ranking, precision, recall, etc., but is not completely determined by them. To argue our thesis, we catalog the Delphic costs and benefits and show how the development of web search over the last quarter century, from its classic Information Retrieval roots to the integration of large Language Models and Generative AI, was driven to a great extent by the quest to decrease Delphic costs and increase Delphicbenefits.

We hope that the Delphic costs framework will engender new ideas and new research for evaluating and improving the web experience for everyone.

This talk reflects joint work with Preston McAfee and MarcNajork.


<\/p><\/div><\/div>\n\n\n\n

The Journey to A Knowledgeable Assistant Leveraging LLMs<\/mark><\/h3>\n\n\n\n
<\/div><\/div><\/div>\n\n\n\n

Abstract:<\/strong>
Large Language Models (LLMs) have demonstrated strong capabilities in comprehending and generating human language, as well as emerging abilities like reasoning and using tools. These advancements have been revolutionizing techniques on every front, including the development of personal assistants. However, their inherentlimitations such as lack of factuality and hallucinations make LLMsless suitable for creating knowledgeable and trustworthy assistants.

In this talk, we describe our journey in building a knowledgeable AI assistant by harnessing LLM techniques. We start with acomprehensive set of experiments designed to answer the questionsof how reliable are LLMs on answering factual questions andhow the performance differs across different types of factual knowledge. Subsequently, we constructed a federated Retrieval-AugmentedGeneration (RAG) system that integrates external information from both the web and knowledge graphs in text generation. This system supports conversation functionality for the Ray-ban Meta smartglasses, providing trustworthy information on real-time topics like stocks and sports, and information on torso-to-tail entities such as local restaurants. Additionally, we are exploring the potential of external knowledge to facilitate multi-modal Q&A. We will share our techniques, our findings, and the path forward in this talk.
<\/p><\/div><\/div>\n\n\n\n

<\/path><\/svg><\/span><\/div><\/div><\/section>\n\n\n\n

Schedule<\/mark><\/h2>\n\n\n\n
March 4, 2023<\/strong><\/td>Monday<\/strong><\/td>WSDM Industry Day Schedule<\/strong><\/td><\/td><\/td><\/td><\/td><\/tr>
Start time<\/strong><\/td>Duration<\/strong><\/td>End Time<\/strong><\/td>Description<\/strong><\/td>Authors<\/strong><\/td>Session (Chair)<\/strong><\/td>Session Topic (paper IDs)<\/strong><\/td><\/tr>
08:30<\/td>01:00<\/td>09:30<\/td>Keynote talk<\/strong><\/td>Andrei Broder (Google Research)<\/td>1 (Sohini)<\/td>Delphic Costs and Benefits in Web Search: A utilitarian and historical analysis<\/td><\/tr>
09:30<\/td>00:15<\/td>09:45<\/td>Contributed<\/td>Sagar Goyal and Eti Rastogi<\/td><\/td>HealAI: A Healthcare LLM for Effective Medical Documentation<\/td><\/tr>
09:45<\/td>00:15<\/td>10:00<\/td>Contributed<\/td>Mukkamala Venkata Sai Prakash, Ganesh Parab, Meghana Veeramalla, Siddartha Reddy, Varun V, Saisubramaniam Gopalakrishnan, Vishal Pagidipally and Vishal Vaddina<\/td><\/td>Accelerating Pharmacovigilance using Large Language Models<\/td><\/tr>
10:00<\/mark><\/td>00:30<\/mark><\/td>10:30<\/mark><\/td>Coffee Break<\/mark><\/td><\/td><\/td><\/td><\/tr>
10:30<\/td>00:30<\/td>11:00<\/td>Invited talk<\/strong><\/td>Haixun Wang (Instacart)<\/td>2 (Sohini)<\/td>LLMs for E-Commerce<\/td><\/tr>
11:00<\/td>00:15<\/td>11:15<\/td>Contributed<\/td>Sudeep Das, Raghav Saboo, Chaitanya S. K. Vadrevu, Bruce Wang and Steven Xu<\/td><\/td>Applications of LLMs in E-Commerce Search and Product Knowledge Graph: the DoorDash Case Study<\/td><\/tr>
11:15<\/td>00:15<\/td>11:30<\/td>Contributed<\/td>Hrishikesh Ganu, Saikat Kumar Das, Akhil Raj, R Sandeep, Satyajeet Singh and Sreekanth Vempati<\/td><\/td>Maya – a Conversational Shopping Assistant for Fashion at Myntra<\/td><\/tr>
11:30<\/td>00:15<\/td>11:45<\/td>Contributed<\/td>Contributed
Harikrishnan C, Giridhar Murthy and Kumar Rangarajan<\/td>		<\/td>Augmenting Keyword-based Search in Mobile Applications using LLMs<\/td><\/tr>
11:45<\/td>00:15<\/td>12:00<\/td>Contributed<\/td>Sachin Farfade, Sachin Vernekar, Vineet Chaoji and Rajdeep Mukherjee<\/td><\/td>Scaling Use-case Based Shopping using LLMs<\/td><\/tr>
12:00<\/mark><\/td>01:30<\/mark><\/td>13:30<\/mark><\/td>Lunch<\/mark><\/td><\/td><\/td><\/td><\/tr>
13:30<\/td>01:00<\/td>14:30<\/td>Keynote talk<\/strong><\/td>Luna Dong (Meta)<\/td>3 (Kumar)<\/td>The Journey to A Knowledgeable Assistant Leveraging<\/td><\/tr>
14:30<\/td>00:15<\/td>14:45<\/td>Contributed<\/td>Wei Qiao, Tushar Dogra, Otilia Stretcu, Yu-Han Lyu, Tiantian Fang, Dongjin Kwon, Chun-Ta Lu, Enming Luo, Yuan Wang, Chih-Chun Chia, Ariel Fuxman, Ranjay Krishna and Mehmet Tek<\/td><\/td>Scaling Up LLM Reviews for Google Ads Content Moderation<\/td><\/tr>
14:45<\/td>00:15<\/td>15:00<\/td>Contributed<\/td>Sohini Roychowdhury<\/td><\/td>Journey of Hallucination-minimized Generative AI Solutions for Financial Decision Makers<\/td><\/tr>
15:00<\/mark><\/td>01:00<\/mark><\/td>16:00<\/mark><\/td>Coffee Break<\/mark><\/td><\/td><\/td><\/td><\/tr>
16:00<\/td>00:15<\/td>16:15<\/td>Contributed<\/td>Jianling Wang, Haokai Lu and Minmin Chen<\/td>4 (Sohini)<\/td>Fresh Content Recommendation at Scale: A Multi-funnel Solution and the Potential of LLMs<\/td><\/tr>
16:15<\/td>00:15<\/td>16:30<\/td>Contributed<\/td>Shreya Saxena, Siva Prasad, I Muneeswaran, Advaith Shankar, V Varun, Saisubramaniam Gopalakrishnan and Vishal Vaddina<\/td><\/td>Automated Tailoring of Large Language Models for Industry-Specific Downstream Tasks<\/td><\/tr>
16:30<\/td>00:15<\/td>16:45<\/td>Contributed<\/td>Contributed
Akshay Jagatap and Sachin Farfade<\/td>		<\/td>Recent Advances in Refinement Recommendations<\/td><\/tr>
16:45<\/td>01:15<\/td>18:00<\/td>Poster Session<\/td>All accepted papers<\/td>5 (Kumar)<\/td>Poster Session<\/td><\/tr><\/tbody><\/table><\/figure>\n","protected":false},"excerpt":{"rendered":"

Keynote speakers Delphic Costs and Benefits in Web Search: A utilitarian and historical analysis The Journey to A Knowledgeable Assistant Leveraging LLMs Schedule March 4, 2023 Monday WSDM Industry Day Schedule Start time Duration End Time Description Authors Session (Chair) Session Topic (paper IDs) 08:30 01:00 09:30 Keynote talk Andrei Broder (Google Research) 1 (Sohini) […]<\/p>\n","protected":false},"author":1,"featured_media":0,"parent":0,"menu_order":0,"comment_status":"closed","ping_status":"closed","template":"","meta":{"footnotes":""},"class_list":["post-934","page","type-page","status-publish","hentry"],"_links":{"self":[{"href":"https:\/\/www.wsdm-conference.org\/2024\/wp-json\/wp\/v2\/pages\/934","targetHints":{"allow":["GET"]}}],"collection":[{"href":"https:\/\/www.wsdm-conference.org\/2024\/wp-json\/wp\/v2\/pages"}],"about":[{"href":"https:\/\/www.wsdm-conference.org\/2024\/wp-json\/wp\/v2\/types\/page"}],"author":[{"embeddable":true,"href":"https:\/\/www.wsdm-conference.org\/2024\/wp-json\/wp\/v2\/users\/1"}],"replies":[{"embeddable":true,"href":"https:\/\/www.wsdm-conference.org\/2024\/wp-json\/wp\/v2\/comments?post=934"}],"version-history":[{"count":25,"href":"https:\/\/www.wsdm-conference.org\/2024\/wp-json\/wp\/v2\/pages\/934\/revisions"}],"predecessor-version":[{"id":1218,"href":"https:\/\/www.wsdm-conference.org\/2024\/wp-json\/wp\/v2\/pages\/934\/revisions\/1218"}],"wp:attachment":[{"href":"https:\/\/www.wsdm-conference.org\/2024\/wp-json\/wp\/v2\/media?parent=934"}],"curies":[{"name":"wp","href":"https:\/\/api.w.org\/{rel}","templated":true}]}}