AI析友口味 助尋「隱世食店」

坊間餐廳指南或食評網站眾多,按主流口味推薦,冷門或新開業小店可能被埋沒。中大工程學院團隊針對以上「盲點」,研發出人工智能(AI)餐飲推薦系統,在傳統演算上,系統結合用戶社交平台中互動率高、口味相似的朋友意見作出分析,找出「朋友圈」中「讚好」食店,提供更貼近用戶個性的推薦。有關系統已製成手機應用程式並上架予大眾免費下載。

Date: 
Friday, April 20, 2018
Media: 
Wen Wei Po

分析好友飲食帖文 中大AI餐飲推薦系統獲獎

不少人出外用膳時會參考網上食評的資料和評分,但亦難免中伏。中大工程學院研發一款名為INCOMIRS的人工智能餐飲推薦系統,透過分析用戶在社交平台的好友有關餐飲消費的帖文,再向用戶推介合適餐廳。該項目於第三屆全國青年人工智能創新創業大會中獲得創新類別一等獎。

Date: 
Friday, April 20, 2018
Media: 
Headline Daily

New prosthetic arm from Hong Kong university helps make amputee’s musical dreams come true

Always envious of good musicians, the 37-year-old can now count himself among their number.  A week after receiving a custom-built artificial limb from a team from Chinese University and Prince of Wales Hospital in Sha Tin, he is playing his favourite Canto-pop classic Under the Lion Rock and music by Japanese composer Joe Hisaishi on the violin.

Date: 
Tuesday, April 17, 2018
Media: 
SCMP

截左手無阻小提琴夢 義肢勇士奏出生命樂章

2014年,阿富更成為首屆「今生不做機械人夢想計劃」的得獎者之一,首本名曲是電影《天空之城》主題曲。但他指舊義肢有只可拉三分一弓等缺點,威院矯形及義肢部門得悉情況後,中大機械與自動化工程學系助理教授劉達銘遂聯同同系的研究助理4年級生陳守健(Samuel)及4年級生葉昆宜(亞宜),於課餘時義務改良阿富的義肢。

Date: 
Tuesday, April 17, 2018
Media: 
Sky Post

中大師生特製義肢 截肢漢自如揮琴弓

一個正在上學,一個正在上班,兩名男生因為一名傷殘人士而結緣,合作製造「特別的義肢」。為了助傷殘人士圓音樂夢,即使兩人不懂得拉奏小提琴,他們上網看了又看相關影片,設計圖改了又改,反反覆覆測試製成品,只是希望製作出最合適對象所用,專門用來拉小提琴的義肢。

Date: 
Tuesday, April 17, 2018
Media: 
Ming Pao Daily News

中大生義助截肢人士圓小提琴夢:不想「消費」傷殘 而是追求進步

四肢健全人士學習樂器,有時都會遇上困難,如果是截肢傷殘人士又會如何?岑幸富(阿富)因交通意外導致左手截肢,三年前在友人的感染下,開始冒起學習拉小提琴的念頭。
在醫院的協助下,阿富裝上一隻可以夾上琴弓的義肢,但由於其設計十分簡單,初期只能拉到約四分之一的弓幅,因而侷限阿富的發揮。最近,中大機械與自動化工程學系,為阿富研製拉弓角度更佳的義肢,如今已能拉四分之三的弓幅。
Date: 
Tuesday, April 17, 2018
Media: 
HK01
Name: 
YIP Kim Fung
Title ( post ): 
Senior Lecturer
Department: 
Electronic Engineering
email: 
kfyip [at] cuhk.edu.hk
phone: 
3943 0871
Avatar: 
Class: 
faculty_member
glossary_index: 
Y
Name: 
HO Sin Cheung
Title ( post ): 
Senior Lecturer
Department: 
Systems Engineering and Engineering Management
email: 
sinho [at] se.cuhk.edu.hk
phone: 
3943 8389
website: 
https://sites.google.com/site/drsinho/
Avatar: 
Class: 
faculty_member
glossary_index: 
H

Two Innovative Projects Win Awards from FinTech Hackathons

Date: 
2018-04-11
Thumbnail: 
Body: 

The rapid growth of science and technology boosts financial services around the globe.  Teams from Department of Systems Engineering and Engineering Management have won the Championship in the Bank of China (Hong Kong) (BOCHK) FinTech Hackathon and received First runner-up in the UHackFin, organised by the Hong Kong University of Science and Technology (HKUST).  The team formed by Dr. Gabriel Fung and Dr. Keith Wong from the Department of Systems Engineering and Engineering Management proposed a new platform named ‘A.I. Stock Analyser’ that digests huge amounts of market information and provides personalised stock analysis to meet the rising demands of investors.  The other team, ‘Expeditioner’, formed by students, focused on the past performance of initial coin offering (ICO) for improving the future of financial services. 

A.I. Stock Analyser

The team ‘Lab Viso’, included Dr. Gabriel Fung, lecturer of SEEM and research fellow of the Key Laboratory on High Confidence Software Technologies (Sub-Lab, CUHK), Ministry of Education, and Dr. Keith Wong, lecturer of SEEM, won the Championship in the ‘FinTech Group’ of the BOCHK FinTech Hackathon in March 2018.  The challenge theme was ‘Future Bank and AI’.  Participants were required to propose a cutting-edge investment application to equip investors with customised analysis and information, and seize investment opportunities according to the customers’ risk attitudes, financial situation and investment experience. 

‘The analysis from current online stock investment platforms are quite limited. They only provide general information of a stock in the market, such as the overall trading volume, the number of investors, etc., but lack of customised information, such as the investment preference of people with similar risk appetite on particular stock. Our proposed solution uses artificial intelligence to analyse the financial market in real-time. It enables banks or financial institutions to launch personalised investment advices to their investors so as to improve service level, bring better investment opportunities for investors and bring revolutionary innovation to fintech. This will further enhance the competitiveness of Hong Kong's financial industry,’ said Dr. Gabriel Fung. In addition, there has been discrepancy in analysis and forecasting between different stockbrokers. The team has therefore addressed the discrepancy and established a new platform to scientifically analyse the most recent five-year broker reports using big data, machine learning and natural language processing (NLP).  The platform can also customise investment solutions for every customer, based on their own risk appetite and investment experiences. This way, customers will not have to spend a long time observing the daily trading volume and the huge amount of inconsistent market information. With this assistive analyser, they are able to make investment decisions rapidly and efficiently. The winning team just applied Cyberport Incubation Programme, targeting to launch this application in the near future. 

A.I. Evaluation of ICO

Another team, ‘Expeditioner’, composed of four engineering students, received the First runner-up in the UHackFin – a FinTech Hackathon organised by HKUST last year. There were a total of 27 teams competing.  Participants were required to propose an innovative solution for improving financial services in the future, within 24 hours.  The CUHK team proposed a new platform ‘Icovisor’ for evaluating the performance of initial coin offering (ICO). ICO is a fundraising mechanism offered to investors in exchange for cryptocurrency, such as bitcoin or ethereum, to fund projects for development and operation. Its concept is similar to an initial public offering (IPO), which means issuing shares for financing from the stock market. However, investors easily suffered losses because of insufficient understanding of ICO and the absence of clear guidelines from the financial regulatory bodies and legislation. 

In view of this, ‘Icovisor’ is designed to generate detailed analysis regarding the past performance of each ICO to help investors make wiser choices.  The use of the latest technologies like big data, machine learning and natural language processing (NLP) has been recognised by the judging panel. Given the high potential for its application, the winning team is planning to apply for the Cyberport Creative Micro Fund to further develop this proposal.

 

 

Filter: Dept: 
Faculty
CSE
SEEM
Media Release

Pages