誰在打造 SOMIN
資深行銷人與一位 AI 教授，共同打造一款產品
Aleks Farseev 教授在新加坡國立大學攻讀博士學位，研究同時從多個社群網路中學習使用者輪廓，之後共同創立了 SOMIN。他是數位行銷 AI 領域的研究教授，其在受眾剖析、個性驅動推薦和廣告可解釋 AI 方面的工作，在 Semantic Scholar 上被引用 577 次（截至 2026-10-17）。
除了研究團隊，SOMIN 的行銷人員還把這些模型轉換為品牌和代理商團隊每週都在使用的行動手冊：行銷空白點圖譜、人物誌與消費者矛盾、預先評分概念和內容日曆。
研究者主頁
引用資料來自 Semantic Scholar，2026-10-17。
從論文到產品
如今在 SOMIN 中執行的七條研究主線
下面的每一條研究主線，都對應平台中的一項能力。點擊一篇論文，即可在出版方網站上閱讀。
01
基於多種資料來源的受眾剖析
如何同時利用多個網路上的貼文、圖片、簽到和感測器資料來了解受眾是誰，以及資料來源為何會改變模型學到的內容。
- AAAI 2020 I Know Where You Are Coming From: On the Impact of Social Media Sources on AI Model Performance (Student Abstract)
- AAAI 2017 TweetFit: Fusing Multiple Social Media and Sensor Data for Wellness Profile Learning
- SIGIR 2017 Cross-Domain Recommendation via Clustering on Multi-Layer Graphs
- ACM TOIS 2017 Tweet Can Be Fit: Integrating Data from Wearable Sensors and Multiple Social Networks for Wellness Profile Learning
02
個性驅動的內容
根據受眾的內容推斷其個性，並據此選擇和生成該受眾會有回應的貼文。
- ACM MM 2022 Personality-Driven Social Multimedia Content Recommendation
- Frontiers in Big Data 2022 Do we behave differently on Twitter and Facebook: Multi-view social network user personality profiling for content recommendation
- WSDM 2021 SoMin.ai: Personality-Driven Content Generation Platform
- ICMR workshop 2021 Two-Faced Humans on Twitter and Facebook: Harvesting Social Multimedia for Human Personality Profiling
03
發布前為創意評分
根據廣告創意的文案、圖片、定向和設定預測其表現，並解釋評分，讓行銷人員能夠據此行動。
- ACM TOIS 2025 Fusing Predictive and Large Language Models for Actionable Recommendations in Creative Marketing
- ACM MM 2023 Against Opacity: Explainable AI and Large Language Models for Effective Digital Advertising
- WSDM 2023 SoCraft: Advertiser-level Predictive Scoring for Creative Performance on Meta
04
用於競品分析的可解釋 AI 與大型語言模型
將效果模型與大型語言模型結合，解讀競爭對手的傳播內容，並說明某個做法為什麼有效，而不僅僅是它有效。
- WWW 2025 SOEX: Explainable AI for Competitor Communications Analytics
- ACM MM 2025 SOMIN: An Explainable AI and LLM Platform for Real-Time, Data-Driven Digital Marketing Strategy
- arXiv 2024 SOMONITOR: Combining Explainable AI & Large Language Models for Marketing Analytics
- ACM MM 2016 bBridge: A Big Data Platform for Social Multimedia Analytics
05
代理型 AI 與策略共創
在撰寫之前先檢索真實情境資料的代理，讓內容保持獨特，而不是趨同於相同的泛泛輸出。
- SIGIR 2026 SOMIN: Agentic AI for Automating Professional Visibility and Countering Content Homogenization
- ACM MM 2025 MindFuse: Towards GenAI Explainability in Marketing Strategy Co-Creation
06
創作者與網紅分析
找到合適的創作者和帳號，用更少的標註對社群貼文進行分類，並區分個人帳號、品牌帳號和名人帳號。
- ACM MM workshop 2021 Improving Multimodal Data Labeling with Deep Active Learning for Post Classification in Social Networks
- AAAI 2019 A Whole New Ball Game: Harvesting Game Data for Player Profiling
- ACM MM 2018 SoMin.ai: Social Multimedia Influencer Discovery Marketplace
- EGOSE 2018 Person, Organization, or Personage: Towards User Account Type Prediction in Microblogs
出版物
頂級會議和期刊論文
發表於 AAAI、ACM Multimedia、SIGIR，以及 ACM TOIS 和 ACM TIST 的完整論文和期刊文章。
- ACM TOIS · 2025 · 期刊文章Fusing Predictive and Large Language Models for Actionable Recommendations in Creative MarketingQi Yang, Aleksandr Farseev, Marlo Ongpin, Alfred Huang, et al.. ACM Transactions on Information Systems, vol. 43, pp. 1-31.
- ACM MM · 2022 · 完整論文Personality-Driven Social Multimedia Content RecommendationQi Yang, Sergey Nikolenko, Alfred Huang, Aleksandr Farseev. Proceedings of the 30th ACM International Conference on Multimedia (MM '22), Lisboa, pp. 7290-7299.開放取得 PDF
- AAAI · 2017 · 完整論文TweetFit: Fusing Multiple Social Media and Sensor Data for Wellness Profile LearningAleksandr Farseev, Tat-Seng Chua. Thirty-First AAAI Conference on Artificial Intelligence (AAAI-17), Technical Track: AI and the Web.開放取得 PDF
- SIGIR · 2017 · 完整論文Cross-Domain Recommendation via Clustering on Multi-Layer GraphsAleksandr Farseev, Ivan Samborskii, Andrey Filchenkov, Tat-Seng Chua. Proceedings of the 40th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR '17), Tokyo, pp. 195-204.
- ACM TOIS · 2017 · 期刊文章Tweet Can Be Fit: Integrating Data from Wearable Sensors and Multiple Social Networks for Wellness Profile LearningAleksandr Farseev, Tat-Seng Chua. ACM Transactions on Information Systems, vol. 35, no. 4, pp. 1-34.
- ACM TIST · 2017 · 期刊文章Learning User Attributes via Mobile Social Multimedia AnalyticsLiqiang Nie, Luming Zhang, Meng Wang, Richang Hong, Aleksandr Farseev, Tat-Seng Chua. ACM Transactions on Intelligent Systems and Technology, vol. 8, no. 3, pp. 1-19.
頂級會議上的論文、示範和摘要
在 ACM Multimedia、AAAI、SIGIR、WSDM、The Web Conference、ICMR、CLEF 和 WebSci 上發表的論文、產業與示範論文以及學生摘要。
- SIGIR · 2026 · 論文SOMIN: Agentic AI for Automating Professional Visibility and Countering Content HomogenizationAleksandr Farseev, Kirill Lepikhin, Kevin Manuel, Maksim Gorodilov, et al.. Proceedings of the 49th International ACM SIGIR Conference on Research and Development in Information Retrieval (SIGIR '26).
- ACM MM · 2025 · 論文MindFuse: Towards GenAI Explainability in Marketing Strategy Co-CreationAleksandr Farseev, Marlo Ongpin, Qi Yang, Ilia Gossoudarev, et al.. Proceedings of the 33rd ACM International Conference on Multimedia (MM '25), Dublin.開放取得 PDF
- WWW · 2025 · 配套論文SOEX: Explainable AI for Competitor Communications AnalyticsAleks Farseev, Sergey Nikolenko, Qi Yang, Kirill Lepikhin, et al.. Companion Proceedings of the ACM Web Conference 2025 (WWW '25), Sydney, pp. 2823-2826.
- ACM MM · 2025 · 產業展示SOMIN: An Explainable AI and LLM Platform for Real-Time, Data-Driven Digital Marketing StrategyAleksandr Farseev. Proceedings of the 33rd ACM International Conference on Multimedia (MM '25), Dublin, industry demonstration.
- ACM MM · 2023 · 論文Against Opacity: Explainable AI and Large Language Models for Effective Digital AdvertisingQi Yang, Marlo Ongpin, Sergey Nikolenko, Alfred Huang, Aleksandr Farseev. Proceedings of the 31st ACM International Conference on Multimedia (MM '23), Ottawa, pp. 9299-9305.開放取得 PDF
- WSDM · 2023 · 示範論文SoCraft: Advertiser-level Predictive Scoring for Creative Performance on MetaAlfred Huang, Qi Yang, Sergey Nikolenko, Marlo Ongpin, et al.. Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining (WSDM '23), Singapore, demonstration.
- WSDM · 2023 · 示範論文"Just To See You Smile": SMILEY, a Voice-Guided GUY GANQi Yang, Christos Tzelepis, Sergey Nikolenko, Ioannis Patras, Aleksandr Farseev. Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining (WSDM '23), Singapore, demonstration.
- WSDM · 2021 · 示範論文SoMin.ai: Personality-Driven Content Generation PlatformAleksandr Farseev, Qi Yang, Andrey Filchenkov, Kirill Lepikhin, et al.. Proceedings of the 14th ACM International Conference on Web Search and Data Mining (WSDM '21), demonstration.開放取得 PDF
- AAAI · 2020 · 學生摘要I Know Where You Are Coming From: On the Impact of Social Media Sources on AI Model Performance (Student Abstract)Qi Yang, Aleksandr Farseev, Andrey Filchenkov. Thirty-Fourth AAAI Conference on Artificial Intelligence (AAAI-20), Student Abstract Track.開放取得 PDF
- AAAI · 2019 · 學生摘要A Whole New Ball Game: Harvesting Game Data for Player ProfilingIvan Samborskii, Aleksandr Farseev, Andrey Filchenkov, Tat-Seng Chua. Thirty-Third AAAI Conference on Artificial Intelligence (AAAI-19), Student Abstract Track, pp. 10025-10026.開放取得 PDF
- ACM MM · 2018 · 論文SoMin.ai: Social Multimedia Influencer Discovery MarketplaceAleksandr Farseev, Kirill Lepikhin, Hendrik Schwartz, Eu Khoon Ang, Kenny Powar. Proceedings of the 26th ACM International Conference on Multimedia (MM '18), Seoul, pp. 1234-1236.
- CLEF · 2018 · 論文Multi-view Personality Profiling Based on Longitudinal DataKseniya Buraya, Aleksandr Farseev, Andrey Filchenkov. Experimental IR Meets Multilinguality, Multimodality, and Interaction: CLEF 2018, Lecture Notes in Computer Science vol. 11018, pp. 15-27.
- AAAI · 2017 · 學生摘要Towards User Personality Profiling from Multiple Social NetworksKseniya Buraya, Aleksandr Farseev, Andrey Filchenkov, Tat-Seng Chua. Thirty-First AAAI Conference on Artificial Intelligence (AAAI-17), Student Abstract Track.開放取得 PDF
- ACM MM · 2016 · 示範論文bBridge: A Big Data Platform for Social Multimedia AnalyticsAleksandr Farseev, Ivan Samborskii, Tat-Seng Chua. Proceedings of the 24th ACM International Conference on Multimedia (MM '16), Amsterdam, pp. 759-761.
- ICMR · 2015 · 完整論文Harvesting Multiple Sources for User Profile Learning: a Big Data StudyAleksandr Farseev, Liqiang Nie, Mohammad Akbari, Tat-Seng Chua. Proceedings of the 5th ACM International Conference on Multimedia Retrieval (ICMR '15), Shanghai, pp. 235-242.
- WebSci · 2015 · 論文Cross-Social Network Collaborative RecommendationAleksandr Farseev, Denis Kotkov, Alexander Semenov, Jari Veijalainen, Tat-Seng Chua. Proceedings of the ACM Web Science Conference (WebSci '15).
期刊、研討會與預印本
- arXiv · 2024 · 預印本SOMONITOR: Combining Explainable AI & Large Language Models for Marketing AnalyticsAleksandr Farseev, Qi Yang, Marlo Ongpin, Ilia Gossoudarev, et al.. arXiv preprint arXiv:2407.13117（未經同行評審）。
- WSDM · 2023 · 演講摘要Under the Hood of Social Media Advertising: How Do We use AI Responsibly for Advertising Targeting and Creative EvaluationAleksandr Farseev. Proceedings of the Sixteenth ACM International Conference on Web Search and Data Mining (WSDM '23), Singapore（演講摘要）。
- Frontiers in Big Data · 2022 · 期刊論文Do we behave differently on Twitter and Facebook: Multi-view social network user personality profiling for content recommendationQi Yang, Aleksandr Farseev, Sergey Nikolenko, Andrey Filchenkov. Frontiers in Big Data, vol. 5 (Original Research).開放取得 PDF
- ICMR workshop · 2021 · 研討會論文Two-Faced Humans on Twitter and Facebook: Harvesting Social Multimedia for Human Personality ProfilingQi Yang, Aleksandr Farseev, Andrey Filchenkov. Proceedings of the 2021 ACM Workshop on Intelligent Cross-Data Analysis and Retrieval (ICDAR @ ICMR '21).開放取得 PDF
- ACM MM workshop · 2021 · 研討會論文Improving Multimodal Data Labeling with Deep Active Learning for Post Classification in Social NetworksD. Krylov, S. Poliakov, N. Khanzhina, A. Zabashta, A. Filchenkov, Aleksandr Farseev. Workshop on Multimedia Understanding with Less Labeling (MULL @ ACM Multimedia 2021).
- EGOSE · 2018 · 論文Person, Organization, or Personage: Towards User Account Type Prediction in MicroblogsIvan Samborskii, Andrey Filchenkov, Georgiy Korneev, Aleksandr Farseev. International Conference on Electronic Governance and Open Society (EGOSE 2018).
- SIGWEB Newsletter · 2016 · 文章360° user profiling: past, future, and applicationsAleksandr Farseev, Mohammad Akbari, Ivan Samborskii, Tat-Seng Chua. ACM SIGWEB Newsletter, 2016, pp. 1-11.
演講與講座
主題演講、講座與 Marketing Mondays 專場
Aleks Farseev 教授談廣告中的 AI，從 ACM Multimedia、INSEAD 到 Marketing Mondays。影片將在 YouTube 開啟。
Pitch · 2026From 6 Weeks to 13 Minutes: Pitching SOMIN AI at ATxSG 2026ATxSG 2026 (Asia Tech x Singapore)；Marketing Mondays by SOMIN
SOMIN 四大模組（競品洞察、空白點發現、概念生成、上市前廣告評分）的新創 Pitch，並附 Mothercare 案例研究。
產業專家演講 · 2025Will AI Make Agencies Obsolete? Rethinking the Future of AdvertisingACM Multimedia 2025，都柏林（Industry Expert Talk IET #2）；影片見 Marketing Mondays by SOMIN對比 AI 進入廣告業的三種方式（取代、輔助和解釋創意工作），以及廣告代理商如何適應。
Pitch · 2025SOMIN @madfestlondon 2025 - L'Oreal Brand Challenge CompetitionMAD//Fest London 2025；Marketing Mondays by SOMINSOMIN 參加 L'Oreal 品牌挑戰賽的作品，主題為可規模化、由創作者主導的內容策略（影片發布於 2025-07-14）。
主題演講 · 2025The Future of Generative AI in Marketing CommunicationsDigital Marketing Conference Kuala Lumpur 2025 (DMCKL2025)；Marketing Mondays by SOMIN關於生成式 AI 用於受眾細分和行銷活動搭建的主題演講，並以 Miraflora Wagyu 案例研究結合 SOMONITOR 和 GWI 進行講解。
講座 · 2025INSEAD: How AI Cut Marketing Costs for a Beef BrandINSEAD 講座；Marketing Mondays by SOMIN講座介紹 AI 消費者洞察和競品洞察如何幫助一家美國高級牛肉品牌降低獲客成本。
講座 · 2025NUS Engineering: Business Strategy with GA4, Meta Ads & SOMONITORNUS Engineering（商業工程專業學生）；Marketing Mondays by SOMIN課堂環節，涵蓋 Google Analytics 4、Meta Ads Manager 以及使用 SOMONITOR 進行的競品洞察。
Marketing Mondays · 2024Marketing Mondays: The Significance Of Responsible AI for Marketing Strategy & BeyondMarketing Mondays at Block71 Singapore；Marketing Mondays by SOMIN圍爐談話，探討在行銷中負責任地使用大語言模型，由 Aleks Farseev 教授在 Block71 Singapore 聯合主持。
演講 · 2024The Intelligent Advantage: Large Language Models Redefining Marketing Decision-MakingTechnology for Marketing（YouTube 頻道）大語言模型如何加快競品、人物誌、品牌語調和痛點分析，以及為什麼它們應當輔助而非取代行銷人員。
示範 · 2024SOMONITOR Demo - Travel Agents and Tourist Guides Industry Forum 2024Travel Agents and Tourist Guides Industry Forum 2024（Singapore Tourism Board）；SOMIN 頻道針對旅遊產業的 SOMONITOR 現場示範。
Podcast · 2023 · 俄語«Нужно разобраться со скиллами, с головой, а потом уже деньги искать»（先把技能和思路理清楚，然後再去找錢）MTS StartUp Hub，Podcast《От сида до экзита》（從種子輪到退出）俄語訪談，談及在海外打造 SoMin.ai、風險投資的低迷，以及大語言模型是否真的是成長機會。
示範 · 2023SoMin.ai Platform DemoAlex Farseev（YouTube 頻道）SOADS（長尾廣告定向）和 SOMONITOR（創意評分、SODA 大語言模型助手）模組的示範導覽。
演講 · 2023From Academic Research to Adtech Entrepreneurship: The Journey of Aleks Farseev and Somin.AIDigital Ignition - A Coworking Space人物專訪，講述從學術研究到 Somin.AI 廣告科技平台及其四大模組的歷程。
演講 · 2023Under the Hood of Social Media Advertising: How Do We use AI Responsibly for Advertising Targeting and Creative EvaluationWSDM 2023 (ACM International Conference on Web Search and Data Mining)，新加坡會議演講，介紹廣告定向和創意評分背後的技術，以及如何負責任地使用這些技術。
演講 · 2021Under the Boot of Google and Facebook and How to Crack it for better PerformanceODS AI Global介紹機器學習如何驅動 Google 和 Facebook 廣告平台，以及長尾定向等提升廣告效果的技術。
講座 · 2020Lecture 1: Introduction (ITMO University, Social Media Computing)ITMO University 課程；Aleks Farseev（YouTube 頻道）一門大學課程的開篇講座，介紹社群媒體平台、社群媒體觀測站及其面臨的挑戰。
講座 · 2017360 Degree User Profile Learning from Multiple Social Networks for Wellness and Urban Mobility Applications (Ph.D. defence public seminar)NUS School of Computing博士論文公開答辯，主題是從社群網路和感測器進行多源學習，用於人口特徵、健康和場所推薦任務。
講座 · 2016 · 俄語Профилирование атрибутов пользователей из множества источников данных различной модальности（從多種模態的多個資料來源中分析使用者屬性）Computer Science Club at POMI RAS；Лекториум (Lektorium)俄語講座，介紹如何從多種不同模態的資料來源中推斷使用者屬性。
FAQ
SOMIN 研究常見問題
SOMIN 是建立在真正的 AI 研究之上嗎？
是的。SOMIN 由 Aleks Farseev 教授聯合創辦，其受眾人物誌、創意評分和競品分析背後的模型，均源自他與 SOMIN 團隊自 2015 年以來發表的研究成果，包括 AAAI、ACM Multimedia、SIGIR、WSDM 和 The Web Conference 的論文，以及發表於 ACM Transactions on Information Systems 的論文。
SOMIN 的研究發表在哪裡？
本頁列出 28 篇經同行評審的論文、示範和摘要，發表於 AAAI、ACM MM、ACM TIST、ACM TOIS、ICMR、SIGIR、WSDM、WWW 等會議和期刊。每個條目都連結到出版方頁面或 DOI，如有開放取得 PDF 也一併提供。
誰在打造 SOMIN？
SOMIN 由經驗豐富的行銷人和一位 AI 教授打造。Aleks Farseev 教授負責研究工作並領導公司；SOMIN 團隊既有機器學習研究人員（其中數位是上述論文的合著者），也有每天與品牌和代理商團隊合作的行銷人員。
這些研究如何體現在產品中？
多源受眾畫像分析支撐 SOMIN 的人物誌，個性驅動的推薦塑造與人物誌相匹配的創意，創意成效預測在發布前為概念評分，而可解釋 AI 與大語言模型的結合則解釋競品內容為何有效。代理型 AI 研究透過以真實品類資料為依據，使生成的內容保持獨特。
在哪裡可以觀看關於 SOMIN 的演講？
上方列出了帶連結的演講：ACM Multimedia 2025 的產業專家演講、Digital Marketing Conference Kuala Lumpur 2025 的主題演講、INSEAD 和 NUS 的講座、ATxSG 2026 的 Pitch，以及 YouTube 上的 Marketing Mondays 專場。