State Key Laboratory of Digital Publishing Technology.

The State Key Laboratory of Digital Publishing Technology (abbreviated as DP-Lab) is committed to promoting the technological advancement and applications of digital publishing, under the supervision of the Ministry of Science and Technology of People’s Republic of China, and management of PKU Founder Group Co. Ltd. Through a mode of industry-university-research cooperation, the research outcomes in DP-Lab have been widely applied in publishers, colleges and libraries, with the support of subsidiaries of PKU Founder Group for system implementation and product promotion.
DP-Lab has been contributed to providing key techniques and functions for multiple publishing-related projects and systems, including Intelligent Revision, Analysis of Media and Academic Resources and Understanding of Financial Bills, which help improve the quality of production, distribution and services of publishing resources to a large extent.
To maintain advanced for China’s digital publishing technology, DP-Lab will continuously study cutting-edge techniques which may greatly impact the future development of digital publishing. Meanwhile, DP-Lab will build a public technology platform for promoting the application of research achievements in real-world scenarios.

JiuTian AI R&D Platform

JiuTian AI R&D Platform is the engine of China Mobile 5G+AICDE strategy. It empowers many industries by providing computing power, data, environments and algorithms for AI research and application. JiuTian AI R&D Platform is designed for AI technologies, focusing on network intelligent with the mission of letting intelligence become the original capability of the communication network and supporting the intelligent transformation of business management. The core AI technologies, operation management and engineering projects mesh with each other and jointly provide scaled commercial value. At present, JiuTian AI R&D Platform consists of JiuTian Deep Learning Platform and JiuTian AI Capability Platform, which provide continuous power for AI development and meet the intelligent application requirements of various scenarios.

Tencent AI Lab NLP Research Demos

A list of NLP-related research prototypes by Tencent AI Lab is demonstrated:
1) Text Understanding System: Word segmentation covering rare and fresh phrases, POS tagging with multi-tag-set support, NER with fine-grained entity types and rich semantics.
2) Multimodal digital human: Show a virtual commentator for game Honor of Kings and use it as a showcase to demonstrate the technologies of Tencent AI Lab in multimodal human-computer interaction.
3) Tencent Transmart: A translation assistant where AI technologies are applied to improve the effectiveness and efficiency of human translation in an interactive way.

Navigation Robot(晶灵)、Q-Robot: Smart Service Space(蜂巢空间机器人)、JD Pharm AI(京东AI智慧药学大脑)

JD.com is China’s largest online retailer and its biggest overall retailer. JD.com operates China's largest nationwide logistics network, developed in-house, covering 99% of the population, and is pioneering innovations to transform global retail. JD.com leverages artificial intelligence to enhance several facets of its business, drive growth and foster innovation. JD.com uses computer vision, machine learning, natural language processing, virtual reality, speech recognition and related computational approaches and techniques to enhance its capabilities in e-commerce, smart supply chain, smart logistics and fintech, among other areas.

Virtual Life and Knowledge Graph Solution

Virtual Life products are designed and implemented based on Gowild Artificial Virtual life Engine (GAVE), which endows holographic 3D projection and various characters. Hupo Xuyan（琥珀·虚颜）is the first virtual life which was launched in 2016. Through cameras, infrared ray and face recognition system, HoloEra Hupo Xuyan can recognize the user and make real-time reaction by 3D animation engine. By utilizing knowledge graph and emotion engine, it can better understand the user and provide corresponding services.
For enterprises customers, including government, banks, pharmaceutical companies, real estate developers and hospitals, we also provide consulting and implementing solutions in knowledge graph and intelligent chatbot.

NLP Technologies in Meituan-Dianping

As China's leading e-commerce platform for services, Meituan-Dianping is focusing on its Food + Platform strategy to build a multi-level technology service platform, covering the whole process from demand to supply in people’s daily lives. There are several R&D teams in Meituan-Dianping engaged in NLP technologies. The Meituan Search and NLP Department is the core team of search technology and artificial intelligence technology. The mission of this team is to create the next generation local search engine, leading natural language processing technology and services. This team not only pays attention to the transfer of AI technology, but also conducts long-term cutting-edge research of search, NLP, and knowledge graph. Current projects include the Meituan/Dianping search system, knowledge graph, intelligent customer service, review understanding, and intelligent assistant. The Meituan Voice Interaction Center is responsible for Meituan's speech recognition, semantic understanding, intelligent interaction, knowledge computing and other voice interaction technologies. At present, many intelligent interactive products are applied in Meituan, such as voice assistant, voice robot for customer service, voice calling robot. The Meituan Waimai Data and Intelligence Team is responsible for WaiMai's knowledge graph, semantic search, personalized recommendation, intelligent marketing etc. It helps to improve the conversion rate significantly and as a consequence of it , yield remarkable revenue.

Wribot Chinese Content Creator Engine

Wribot Chinese Content Creator Engine (the Engine) is developed by Beijing Wribot Intelligent Technology Co., Ltd., a company engaged in the Chinese NLP technology. With features like Article Creating, Article Style Changing and Article Posting, the Engine can be used in the creation, distribution and monitoring of various contents. Taking advantage of Wribot’s self-media accounts lineup, the Engine is able to attract massive Internet traffic in a short time, and help improve brand communication while decreasing costs.
The Engine-based products include Article Creator Bot, Content Style Changing & Publishing Bot, and AI Self-media Bots Lineup.
Backed by BlueFocus, the largest communication group in Asia, Wribot has established the technical cooperation with PKU’s NLP academic team, and has gained investments from BlueFocus, Baidu, and Lenovo. Currently, Wribot serves tens of companies among the Fortune 500.

Several advanced NLP technologies and applications from Microsoft are demonstrated.

(1) Microsoft Recommender, an open-source project which aims at democratizing the recommendation technology and filling the gap between academic research and industrial applications. It provides examples and best practices for building recommendation systems, mainly covering five aspects including data preparation, model building, performance evaluation, hyperparameters tuning and Operationalization.
(2) Unicoder and Unicoder-VL, two pre-trained models for cross-lingual and cross-modal scenarios. Unicoder is a universal encoder for multiple languages, which is pre-trained by cross-lingual tasks. With Unicoder, a model can be trained using English training data first (as an NLP task often has rich training data in English) and then directly applied to inputs of the same task in other languages (e.g., French, German, Japanese, Chinese etc.). Unicoder-VL is a universal encoder for vision and language tasks, which is pre-trained by cross-modal tasks. With Unicoder-VL, cross-modal tasks such as visual QA and image-text retrieval are greatly improved. Both Unicoder and Unicoder-VL show strong gains for tasks in Microsoft Bing.
(3) LiveBot, an automatic live commenting system for videos based on deep learning, which simulates human behaviors to comment based on vision and language. (4) Immersive Reader, an extension of machine translation related technologies to let users embed text reading and comprehension capabilities into applications, helping users of any age and reading ability with features like reading aloud, translating languages, and focusing attention through highlighting and other design elements.

Natural Language Processing at Baidu

Baidu has been deeply engaged in the field of natural language processing for many years, focusing on cutting edge NLP research and industrial applications. The booth will show Baidu's latest progress in semantic computing, machine translation, reading comprehension and dialogue. It will also demonstrate Baidu's open NLP services, platforms, and open source projects, including ERNIE, the most effective continual pre-training framework for language understanding, PaddleNLP, the open source NLP toolkit and pre-training model library based on PaddlePaddle, UNIT, the Understanding and Interaction Technology Platform for customized dialogue systems, the Open Translation Platform supporting text, voice, image, and customized translation, the Intelligent Content Creation Platform, and other NLP capacities opened on Baidu Brain.

JoveBird Alternative Data Analysis Platform, JoveEye Global Research and Technology Discovery System, Jove Mind Knowledge Graph, Cognitive Intelligence Technologies

JoveBird: an intelligent big-data platform for financial investment analysis based on professional financial analysis model and leading cross-language big-data technology.
JoveEye: It processes global technology big data in a semantics-based and knowledge-oriented way to perform the functions of knowledge discovery, intelligence mining, technology profiling, and smart evaluation.
Jove Mind Knowledge Graph: a knowledge graph which acquires knowledge automatically from global open-source data and further enables effective transformation from massive data to highly structured, visualized information.
Cognitive Intelligence Technologies: in alliance with world’s foremost universities, research institutes and scientists, GTCOM’s 2020 AI Lab mainly oriented in machine translation, natural language processing, semantic computing, speech recognition, computer vision, semantic computing and intelligent Q&A.

Introduction about Xiaomi

Founded in April 2010, Xiaomi is an Internet company that focuses on smartphones, smart devices and the Internet of Things platform. Xiaomi is dedicated to making amazing products that will touch your heart, and have honest pricing. Committed to continuous innovation on smartphones, smart devices, Internet TV and other products, Xiaomi makes positive adjustments to the enterprise layout, in AI, new retail, internationalization, fintech and so on. With the innovative advantages of its 3-pillar business model: hardware, new retail and Internet, Xiaomi rises rapidly and becomes one of the representatives of innovative “Internet+” companies.
In 2017, only in business for 7 years, Xiaomi grossed over a 100 billion in revenue and broke records in the global business industry. Xiaomi was also listed on the main board of the Hong Kong stock exchange on July 9, 2018 (1810. HK). In 2019, Xiaomi became the youngest company on Fortune Global 500 List and are the fourth Chinese Internet Company being shortlisted.

ByteDance

ByteDance’s platforms enable people to discover and create a world of content powered by technology. We inform, educate, entertain and inspire people across languages, cultures and geographies.

knowledge graph platform & Intelligent Question Answering Platform

KG Data is a high-tech enterprise with knowledge graph as the core technology.The technology covers the whole technology chain of knowledge graph from data acquisition to data cleaning, knowledge extraction, knowledge representation, knowledge reasoning, dynamic ontology modeling, graph storage, visual association analysis, intelligent question and answer, semantic retrieval, etc.Products are widely used in pharmaceutical companies, medical institutions, military institutes, scientific and technological information and publishing and other fields.
The knowledge graph platform has many advantages, such as powerful dynamic ontology modeling, efficient data import, unstructured data processing for human-machine collaboration, full-cycle knowledge map operation and maintenance, intelligent knowledge graph fusion, rapid installation and deployment, and secondary development.
Intelligent Question Answering Platform FAQ and Knowledge Graph Question Answering (KBQA) has the advantages of few data cold start and high accuracy, privatization customization, fast and convenient, safe and efficient, dynamic operation and maintenance closed-loop iteration, self-learning to reduce operation and maintenance costs, can handle a wide range of knowledge, rapid development and deployment.

Multi-Task Multi-Head Attention Memory Network for Fine-Grained Sentiment Analysis

Sentiment analysis is widely applied in personalized recommendation, business reputation monitoring, and consumer-driven product design and quality improvement. Fine-grained sentiment analysis, aimed at directly predicting sentiment polarity for multiple pre-defined fine-grained categories in an end-to-end way without having to identify aspect words, is more flexible and effective for real world applications. Constructing high performance and efficient fine-grained sentiment analysis models requires the effective use of both shared document level features and category-specific features, which most existing multi-task models fail to accomplish. In this paper, we propose an effective multi-task neural network for _ne-grained sentiment analysis, Multi-Task Multi-Head Attention Memory Network (MMAM). To make full use of the shared document level features and category-specific features, our framework adopts a multi-head document attention mechanism as the memory to encode shared document features, and a multi-task attention mechanism to extract category-specific features. Experiments on two Chinese language fine-grained sentiment analysis datasets in the Restaurant-domain and Automotive-domain demonstrate that our model consistently outperforms other compared fine-grained sentiment analysis models. We believe extracting and fully utilizing document level features to establish category-specific features is an effective approach to fine-grained sentiment analysis.

Alibaba Group

Alibaba Group enable businesses to transform the way they market, sell and operate and improve their efficiencies. We provide the technology infrastructure and marketing reach to help merchants, brands and other businesses to leverage the power of new technology to engage with their users and customers and operate in a more efficient way.
During NLPCC 2019, Alibaba Groups will introduce some of the latest technologies in NLP and their applications in our product and scenarios including machine translation, intelligent customer services, smart mobile browser and so on.

NLP data collection and annotation

Since its establishment as an AI data resource provider, Speechocean has always devoted itself to providing specialized engineering data products and services to enterprises and scientific research institutions in the whole industry chain of AI. Our business involves various domains such as speech recognition, speech synthesis, computer vision, lexicon, and natural language processing and provides relevant services for the design, collection, transcription, annotation, etc. of data.

NiuTrans Cloud

NiuTrans Cloud is a machine translation service provider for enterprise users developed by NiuTrans Team with its independently developed neural machine translation engines, supporting 119 languages with Chinese at the core.
NiuTrans Cloud offers services including webAPI, private deployment and custom development of machine translation engines, offline SDK for the mobile smart devices, and a NiuTrans store for some apps, such as mobile apps for Android and iOS, plugins for CAT tools like Trados and memoQ, Office Word and Kingsoft WPS.
NiuTrans Cloud has gained partners of 60+ industrial leaders worldwide like Tencent, iFLYTEK, KingSoft, Aispeech, memoQ and SIPO.