Dec 24, 2020| JD Technology

JD Cloud & AI partners with Beijing Life on Industry Digitalization

Martin Li

by Martin Li

JD Cloud & AI will partner with insurance company Beijing Life to advance digitalization in the insurance industry, based on a cooperation agreement signed by the two parties.

JD Cloud & AI will partner with insurance company Beijing Life to advance digitalization in the insurance industry

The cooperation will regard cloud services, AI, Internet of Things and community health services.

JD Cloud & AI will provide safe and reliable basic cloud services including the public, private and hybrid cloud, helping Beijing Life build its fundamental cloud computing platform.

The application of AI technologies is expected to digitalize the insurance company’s whole business chain.

In addition, JD will use its supply chain to help Beijing Life upgrade its community health service business.

(bjllihao3@jd.com)

Nov 16, 2022| JD Technology

JD Cloud Ranks No.1 in World’s Top Dataset for Task-Oriented Dialogue Modeling

Doris Liu

by Doris Liu

The model proposed by the team of JD Cloud’s intelligent customer service system, Yanxi (言犀), which provides users with 24/7 service in all scenarios, ranked first among models of MultiWOZ, the world’s top dataset for end-to-end task-oriented dialogue modelling.

Mars, an end-to-end task-oriented dialogue system proposed by the Yanxi team, innovatively models the relationship between dialogue context and dialogue/action state through semantic-aware contrastive learning strategies to ensure a better performance of the system. Empirical results show the proposed Mars system maintains state-of-the-art performance on the MultiWOZ 2.0, CamRest676, and CrossWOZ.

Proposed by the University of Cambridge, MultiWOZ, which is short for the Multi-Domain Wizard-of-Oz dataset, is a fully-labeled collection of human-human written conversations spanning over multiple domains and topics. It is one of the most influential task-based dialogue datasets, containing highly natural conversations between tourists and staff obtained from information centers of tourist cities, in which 70 percent of the dialogue is from 2-5 domains in multiple scenarios, and it is becoming one of the most classic and challenging datasets in the field of natural language processing.

Many prestigious universities and research institutions across the world have participated in the competitive challenge since the release of MultiWOZ, including research groups from Tsinghua University, Hong Kong University of Science and Technology, Microsoft Research, Amazon, DeepMind, and Salesforce.

An example of end-to-end task-oriented dialogue generation

As part of the challenge, Yanxi participated in end-to-end task-oriented dialogue generation, which needs to first identify the user’s intention through the user dialogue and generate dialogue states, query the database based on the generated dialogue states for entity matching, and then provide natural language responses that meet the user’s goals according to the dialogue policies based on the query database results.As the industry’s first large-scale commercial intelligent customer service system, Yanxi has accumulated rich practices in task-based dialogue generation through support of the entire chain and lifecycle of JD.com’s customer service, with a generation of 10 million dialogues per day on average, serving JD’s over 580 million customers and 178,000 merchants with a smooth experience.

(liujun215@jd.com)

Nov 9, 2022| JD Technology

JD Technology Sets Mortgage Loan Example with Biological Assets as Collateral

Doris Liu

by Doris Liu

China’s HengFeng Bank, with the assistance of the biological asset supervision platform developed by JD Technology, provided mortgage loans for 160 heads of live beef cattle, the first of such kind across the country, setting an example of digital finance nationwide, as reported by the international market research firm IDC.

According to data from the Chinese Academy of Social Sciences, the funding gap of China’s “three rural issues” – agriculture, rural areas and farmers – reached RMB 3.05 trillion yuan in 2018. However, China’s rural finance is confronted with multiple obstacles, such as insufficiency of effective credit data and lack of collateral. Financial institutions have been reluctant to accept biological assets as collateral, although they hold significant potential as collateral in rural areas, in terms of evaluation, ownership confirmation, supervision, disposal, etc.

Utilizing JD Technology’s platform, HengFeng Bank granted RMB 1.75 million yuan to Wang Yuhe, a cattle producer of Hengchang Farm in Heze, Shandong province. Connecting with camera equipment installed in the cattle ranch and smart ear tags attached to the ears of 160 cattle, the platform collects dynamic data on each head of cattle 24/7 in real time, monitoring and recording the growth information and health status throughout the life cycle of beef cattle production. Meanwhile, the data are synchronously uploaded to blockchain to ensure the uniqueness, authenticity as well as protection against being tampered with or reproduced, largely reducing the bank’s lending risks and easing Wang’s financial pressure.

Creating a digital twin on blockchain ensures that all information of a physical animal will be saved immutably, which enables biological assets to qualify as collateral. The biological asset supervision platform, integrated with AI, big data, blockchain, IoT and more technologies, has been applied in production of livestock including cattle, sheep and pigs, and poultry and fish by monitoring them with cameras, smart IoT devices of ear tags, collars and leg rings; and equipment including water quality sensors, temperature and environment control and ultrasonic instruments to match different animals’ traits.

For example, aquaculture supervision solutions based on the platform were launched in July 2022 by linking dissolved oxygen sensors, pH sensors, Oxidation-Reduction Potential (ORP) sensors and other IoT devices, underwater cameras, intelligent feeding systems and more, which will simplify the process of loan application for fish farmers and banks.

On the consumer side, JD Technology’s platform also guarantees product quality, as it supports full traceability of meat products. In some cases, the amount of walking chickens have done before butchery can be assessed with a specially designed pedometer similar to a poultry leg band, which is connected to the platform, to ensure the quality of the meat.

Farmers can more easily spot any risks to their livestock, plants and the environment with the platform. The IDC report believes that the digitalized biological asset supervision helps bring more financial support to rural areas. Apart from loans, financial institutions can lower the threshold to offer bespoke livestock insurance to mitigate the multitude of risks faced by farmers. Furthermore, governments are able to distribute subsidies to farmers more precisely through the platform, further boosting the ongoing rural revitalization.

(liujun215@jd.com)