The Key Laboratory of Mongolian Information Processing Technology of Inner Mongolian: Focus on Research of Mongolian Information Processing and Development of Intelligent Mongolian Software

The laboratory was established in 2007. After many years of construction and development, the laboratory has independently developed Mongolian information retrieval system, Mongolian speech recognition system, Mongolian text-to-speech system, Mongolian Optical Character Recognition system, Mongolian-Chinese translation system, Cyrillic Mongolian-traditional Mongolian conversion system, and Mongolian input method. Mongolian-Chinese-English Electronic Dictionary, Mongolian ancient literature retrieval system, etc.

In-depth and Multi-aspect Analysis of Contents of Digital Resources

The State Key Laboratory of Digital Publishing Technology (SKL-DP) has produced lots of innovative and practical achievements through in-depth and multi-aspect analysis of digital publication resources. Based on the start-of-the-art techniques of machine learning and natural language processing, we have realized more accurate processing of textual data, more precise recognition of components in a document, and more intelligent typesetting of journal publication and customization of knowledge service. Based on our research outcomes, the online production of journals has come true, the mean average precision (MAP) of recognizing complex components in a document is improved to be larger than 80%, and several knowledge service products such as searching platform of similar legal cases, and Fangzheng Xiaokan smart staff member have been applied in practice. Thus, a complete procedure of processing and analyzing digital publications has been established.

AI-S Selling Assistant, AI-S Customer Service Assistant for Retail, AI-S Customer Service Assistant for Banking, AI-S Big Data Analytic Platform, AI-S Poetic Talent

AI-Strong is the industry leading NLP application company in China which focus on providing intelligent services through up-to-date NLP along with other AI and big data technologies. So far, AI-Strong has developed its own task-oriented dialogue system based on deep learning, and successfully applied the system into applications of pre-sales, intelligent customer services, and quiz show etc. AI-Strong is fully bent on enhancing customer capacity in taking full advantage of their data assets. In our demos，AI-S Selling Assistant is a chatbot selling assistant for cosmetic products for e-commerce platform; AI-S Customer Service Assistant for Retail can automatically handle customer feedback and complain emails, include email content understanding and appropriate reply generating; AI-S Customer Service Assistant for Banking is a chatbot that provide customer services for bank clients through social media; AI-S Big Data Analytic Platform integrates a powerful tool sets for non-structured data collection, parsing ( PDF & OCR recognition tech), NLP+, knowledge graph, data visualization; AI-S Poetic Talent(诗词小强) demonstrates the ability in Chinese Language understanding and generating. Thanks to great support from Industrial Bank, Pingan Bank, VIPShop, Globalgrow Ecommerce, etc. We are advancing forward steadily and fast in fields of AI.

JD Instant Messenger Intelligence(JIMI)

JD Instant Messenger Intelligence(JIMI) is an intelligent customer service system developed by JD AI Platform and Research Group. By applying AI technologies such as natural language processing, deep neural network and machine learning, JIMI can provide shopping advice and after-sale services, and now it can answer most of questions about more than 1 billion products, and undertake over 90% customer services of JD.com.
Compared to other Chatbots, JIMI could both solve problems and recognize sentiments. It not only answers the questions of customers but also tracks customers’ problems until they are completely solved. Besides, it also has sentiment-analysis ability and high EQ, making interaction warmer. JIMI is also the first large-scale commercial intelligent Chatbot with sentiment-analysis in the industry.

Tencent LINGO Lab

Tencent LINGO Lab, is a research team under Tencent's Mobile Internet Group, focusing on NLP application research.
The LINGO Lab is positioned to provide NLP supports for MIG products, including Tencent Mobile Browser, Tencent Desktop Browser, Tencent Map, Tencent Internet of Vehicles, Tencent Autonomous Driving, Tencent Translator, Tencent Dingdang, etc.
The demands of users, overcoming language barriers among human-machine and realizing more convenient human-machine interaction, is ever-increasing. In view of this, the LINGO Lab focuses on technical study and application exploration of the above issue. Its research areas includes deep understanding of texts, machine translation, location-based services, knowledge Q&A, dialogue systems and so on.
The LINGO Lab is committed to developing leading NLP technologies and exploring potential business scenarios that are empowered by NLP. It also plays an active role in connecting with academia and advanced talents.

ABitAI Tech: AI Interactive Dialog System Robota

The first of its kind in China, AbitAI Tech is a leading B2B e-Commerce customer service platform enabled by artificial intelligence (AI) and human-computer interactive. It offers tailored purchase recommendation solutions. By integrating the knowledge graph of a certain vertical domain, natural language understanding(NLU) technology and personalized recommendation, AbitAI Tech has developed a cutting edge AI interactive dialog system ROBOTA to help corporate clients build a deep connection with customers in cost effective manners to achieve high customer recruiting and repeated purchase rates. Currently, our clientele boasts well known brands such as VIP.com, McDonalds, TCL, Chow Tai Fook.

Sogou: From Keywords Search to Question Answering

Instead of scrolling through page after page of search results and trying to figure out which information to trust, consumers have come to expect and demand direct answers to their queries. In that context, Sogou Search tries to think on behalf of humans and has a wide range of innovative products and services that are used by hundreds of millions of users in China. Focusing on natural interaction and knowledge computing, Sogou Search has made significant breakthroughs in voice and image technologies, machine translation, and Q&A, which have been successfully integrated into our products and services. Our proven AI capabilities will facilitate the launch of more disruptive products and services, such as virtual personal assistants, or VPAs, to serve users anytime, anywhere.

AISpeech Co.,Ltd. TalkingGenie

Talking Genie, is a newly launched platform by AIspeech for enterprises to customize their own virtual business assistants, who can provide 7X24hr interactive self-services about key information they want to communicate with their customers. It adopted the innovative Heuristic Conversation Technology and the information management technology of complex-structured knowledge pioneered by AISpeech, In the conversation, the virtual assistant will keep suggesting related topics to the customer and lead the conversation carrying smoothly. With the Talking Genie, enterprises can efficiently manage their own information and knowledge in various structures and convey important information to their customers, and even guide customers’ intention to achieve communication and business goals.

Liulishuo

In July 2016, Liulishuo officially launched its AI English teacher to offer users personalized and systematic English learning solutions. It has equipped 3 self-developed AI technologies.
1) Speech Recognition and Scoring Engine. It’s a multi-dimensional, multi-granularity speech evaluation engine with an offline mode, based on Deep Learning technology.
2) Writing Assessment and Enhancement Engine. By developing capabilities in grammar error detection, pragmatics error detection, synonym analysis and semantic understanding, this engine can evaluate one’s speech based on pronunciation, vocabulary, grammar, fluency and pragmatics, locate errors and provide suggestions for improvement.
3) Adaptive Learning System. The company has built a brand-new teaching model through continuous fast learning and self-evolution like AlphaGo. According to a study using the TOEIC Bridge™ test from ETS, Liulishuo’s AI teacher is proven to triple users’ learning efficiency over human teachers.

China Mobile intelligent customer service

China Mobile intelligent customer service is jointly developed by the China Mobile online Services Company Limited and China Mobile Research Institute. It is currently providing services to more than 200 CMCC internal channels, with a total of 110 million monthly visits, using a variety of technologies such as dialogue, search, and NLP.

Alibaba Group

ALIBABA GROUP’S MISSION IS TO MAKE IT EASY TO DO BUSINESS ANYWHERE. We enable businesses to transform the way they market, sell and operate. We provide the fundamental technology infrastructure and marketing reach to help merchants, brands and other businesses to leverage the power of the Internet to engage with their users and customers. Our businesses are comprised of core commerce, cloud computing, digital media and entertainment, innovation initiatives and others. Through investee affiliates, Cainiao Network and Koubei, respectively, we participate in the logistics and local services sectors. website: http://alibabagroup.com/en/global/home

YeeSight Big-Data Ecosystem, NexMagic Semantic Computing, Cognitive Intelligence Technologies, Jove Mind Knowledge Graph

YeeSight Big-Data Ecosystem: a big-data ecosystem which build on the basis of natural-language processing, neural network machine translation, big-data analysis and artificial intelligence technologies, focusing on big-data platforms, big-data applications and industry solutions, including Insider, Jove Bird, Public Safety, Jove Eye, Cover Global Index (CGI), etc.
NexMagic Semantic Computing: a vertically-oriented knowledge discovery and knowledge service ecosystem which build on the basis of big data, multilingual machine translation, artificial intelligence algorithms and knowledge graph modeling. NexMagic has opened up knowledge services for press and publication, science and technology and enterprises.
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.
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. Thus, Jove Mind is capable of providing knowledge data services as well as providing algorithm services for all professions and industries.

Microsoft Research Asia

Several advanced NLP technologies and applications from Microsoft are demonstrated. The first one is about Question Answering (QA) and Natural Language Understanding (NLU). Various QA systems have been built based on different types of data resources, including knowledge base, web tables, texts, images, and Q-A pairs. Many technologies behind these QA systems are used in Microsoft products like Bing, Cortana and Xiaoice. The semantic parsers for NLU tasks were built based on structured knowledge bases including knowledge base and web tables, which are useful to better understanding human language and design dialogue systems. The second one is about social computing project, which engages in interdisciplinary research on intersecting computer science with social science, psychology and cognitive science. It aims to gain an understanding of users in depth and provide them with intelligent and personalize online experience based on large scale human behavioral data on various heterogeneous social platforms. The technologies have been used in various AI products in Microsoft. The third one is about neural machine translation customization, which allows users to customize Microsoft Translator's advanced neural machine translation with seamlessly integration of their own text data. This released service can bring gains of MT quality varying from a noticeable few BLEU points to a remarkable jump of over 10 BLEU points.

Natural Language Processing at Baidu

Some advanced NLP technologies and applications from Baidu will be demonstrated. We will update Baidu's research advancement and achievement in NLP technologies, such as machine translation, question answering and dialogue. We will also demonstrate the impact of NLP technologies in Baidu's core products, including innovative applications and extensive impacts. Now the NLP technologies support about 200 applications at Baidu and the modules are called up more than 200 billion times per day. Some of them are available at Baidu’s AI platform（http://ai.baidu.com）, including syntactic parsing, semantic analysis and calculation, sentiment analysis, machine translation, understanding and interaction technology etc. And we open some NLP technologies at Github platform（https://github.com/baidu），including lexical analysis，sentiment analysis, Famillia topic model, and AnyQ-FAQ system etc.

Leyan Cognitive Platform

Leyan Cognitive Platform is an integrated platform that empowers knowledge-intensive industries with AI capabilities. Core to the platform is the engine comprised of several fundamental modules, including a Natural Language Processing Module, a Knowledge Graph Module, a Multi-turn Dialog Management Module which is based on adaptive learning, an Intelligent Recommandation Module, and a Machine Learning Module. This breakthrough technology is invented for the aim to resolve the ubiquitous pain points prevailing in the intelligence-intensive work, such as the organization of information from heterogeneous data sources by manual labor and the monotonous repetitive question answering provided by online attendants. Equipped with AI powers, Leyan Cognitive Platform is able to process and understand text of high complexity, and also generate accurate predictions and insights to help users make wiser decisions with much less effort and time. It also proves to be a great success in application compatibility across different industries; custom design products have been launched for the industry of finance, medicine, e-commerce, education, and e-government and are gaining in popularity for its powerful AI functions, user-friendly management suites, and most importantly, the unparalleled ease of use.

An Intelligent Business Messaging and Conversation Solution

Founded in 2015, Laiye is a leading AI interactive platform that empowers consumers and enterprises with intelligent assistant products built with natural language processing, multi-turn dialogue management and context-based recommendation technologies. Its main product, Wulai, which allows enterprises to build their own chatbots, aims to help enterprise customers increase their sales revenue,marketing ROI and service efficiency through a whole package of business messaging and conversation solutions. So far the company has partnered with and been trusted by a variety of clients, such as China Mobile, Wyeth, Friso, McDonald’s, Burger King, Ctrip, Qunar,Meituan, SPD Bank and so on.
Laiye completed its Series B funding invested by Wu Capital, Sequoia and Lightspeed in 2017. This was the third consecutive investment Laiye received from Sequoia and Lightspeed. Before that, Laiye had received a round of financing led by Microsoft, which was the tech giant’s first direct VC investment in Asia.

Huawei: Noah’s Ark Lab

Founded in 2012, Noah’s Ark Lab is the AI research center of Huawei Technologies. We undertake the following primary domains to drive fundamental R&D in Artificial Intelligence:
(1) Speech & Language Processing, including: Speech Recognition and Synthesis, Machine Translation, Dialogue System, Natural Language Generation, Question Answering, Deep Learning for NLP
(2) Computer Vision
(3) Search & Recommendation
(4) Decision & Reasoning
(5) AI Theory
Researchers in the Speech & Language Processing Lab (of Noah’s Ark Lab) have been conducting academic and applied research, and also developing AI-enabled products and services. For example, we have been working with the teams at Huawei consumer business group to build the intelligent voice assistant on Huawei mobile devices. We are developing a series of technologies, such as far-field speech recognition, task-oriented multi-turn dialogue systems, and neural machine translation, to enhance the interaction experiences of Huawei’s smartphones. Some technologies have been shipped to Huawei smartphones since the Mate 10 series release.

Media Dissector & Media Dissector

Media Dissector is our mass media, social media and user-generated content media monitoring and analytics tool. Media Dissector applies distributed cloud computing crawler technology to capture media information on the internet. It uses Chinese Natural Language Processing technology such as word segmentation, named entity recognition and Chinese information processing technology to process information and provides correlation analysis to enable drill downs for information discovery.
Law Dissector is our search and statistical analysis tool for legal research. It utilizes our natural language processing technology and applies data science methodology in the legal industry. Law Dissector summarizes the key attributes of each legal case and aggregates them in a broader statistical scope, which enables judges, lawyers and researchers to find implied patterns or rules that would be valuable for their respective professional objectives.

Call Center Chat-bot: SmartServices, an AI-Powered Customer Service that Understands You

SmartServices is an AI powered solution for customer service industry. By integrating AI-powered customer service agent and real customer representatives, it serves customers anytime, anywhere, in their preferred way and with their own languages. It can solve all problems encountered by customers throughout the entire product lifecycle.
The key technologies behind the solution include at least the following four categories:
A. Deep natural language understanding - a hybrid-mode language-independent NLU engine was developed to precisely understand customers’ utterances and extract key semantic blocks from customers’ input.
B. Policy-adaptive dialog manager – a sophisticated DM was developed based on self-learning and reinforcement learning technologies, as well as probabilistic state managing engine. It can make the most out of customers’ context, and therefore solve customers’ problems with the least rounds in the most natural way.
C. Personalized and flexible natural language generation – this engine can generate the most suitable utterances to help customers, and make them feel happy during the conversation.
D. Large-scale knowledge graph building – the engine was implemented based on knowledge extraction, fusion and reasoning technologies. It not only helps the chat-bot become smarter and more usable, but also helps build the fundamental base for corporate innovation.

Knowledge-Driven Enterprise Intelligence Platform

Using deep learning, NL technology to build a knowledge driven enterprise intelligence platform, to monitor and interpret all signals from such a variety of different players that related to the enterprise, such as clients, customers, suppliers, competitors, distributors, regulators, investors and technology providers. The platform provides cross-lingual and cross-source knowledge extraction, knowledge graph construction and automatics learning, aim to empower enterprises to utilize the value of data and knowledge worldwide in real-time, enable all businesses adapting to the dynamic globalized economy with the help of AI and Big Data technologies in the international market for the enterprise marketing, product design management, competitor analysis, and supply chain management.

Voice-enabled Medical Record system, Kids Educational Accompany Robot, and Voice Interaction Solution for Internet Of Vehicles

1. Voice-enabled Medical Record System:As the leader of intelligent voice interaction in the medical field, Unisound is the first to propose a voice interaction solution in the medical professional field in China. At present, they have entered the market and have launched 30+ representative top-three hospitals. They have become excellent partners of Peking Union Medical College Hospital for two consecutive years.
2. Kids Educational Accompany Robot（UniToy）UniToy has an excellent product experience. For all kinds of children's smart products, just put the UniToy hardware module into the device, you can upgrade to a function early education robot that can interact with children's all-natural voice. It is market lead in line with the demand of the children's market for fast shipping industry.
3. Internet Of Vehicles Intelligent Voice SolutionThe Unisound intelligent car voice interaction solution is suitable for smart rearview mirrors, vehicle-mounted central control, HUD, car boxes, etc., so as to truly realizing human-machine dialogue and free both hands. At present, it has served nearly 100 aftermarket manufacturers and brand owners, ranking first with nearly 70% in the automotive aftermarket.

The NLP algorithms in Aimi

The introduction of product achievements:
Aimi is an intelligent virtual assistant platform developed by Xiao Mi Company, which is launched to the market on July 23, 2017. By the end of March 2018, the total number of activated devices exceeded 23 million. Natural Language Processing algorithms are the foundations of virtual assistant platform serving stacks. Aimi utilizes the following NLP algorithms:
1. The basic level covers knowledge graph, multi-dimension user profile, large scale machine learning, etc.
2. Sentence understanding level covers word segmentation, part-of-speech tagging, chunking analysis, etc.
3. Semantic understanding level covers domain classification, slot filling, intent classification, etc.
4. The dialogue management level is based on co-reference resolution, ellipsis resolution, dialogue status management, etc.
5. The result rendering level is based on natural language generation combined with domain information and personalized information.

MAXHUB——Efficient Conference Platform

MAXHUB conference solution offers all-in-one display, high-precision writing and featuring annotation.It brings you device collaboration, wireless sharing and remote conferencing, to help draw conclusions quicker, make meetings more productive, creative and accountable.

ByteDance

Natural language processing technologies have been widely applied in Bytedance's products. Successful NLP applications include topic modelling for document mining, text classification for click-bait title detection, sentiment analysis from user comments, text matching technologies for article search, syntactic and semantic parsing for query understanding, machine translation, and language generation technologies for ads slogan writing.
1.Xiaomingbot Machine writing: Combining grammar-based generation, machine learning methods, and automatic text summarization, Xiaomingbot, a robot writer, produces sports and financial reports as well as Toutiao Alert for Missing People. The robot writer updates its news almost as fast as live streaming TV, and its articles are as professionally written in a way indistinguishable from those of journalists.
2.Machine Translation at ByteDance: Developed independently by ByteDance, ByteTrans supports translation between English and Chinese, and 10 other languages such as Spanish, French, Vietnamese, Malaysian, Arabic and Japanese etc. into English. It facilitates international creators in content creation and user communication. Now, machine translation has been applied to functions such as search and review, and on other social media apps such as TikTok, and Vigo Video.

Media Analytical Technologies from Wisers AI Lab

Wisers AI Lab is dedicated to the research and development of innovative AI technologies for analyzing diversified real-world open-domain big data, to create and power intelligent enterprise solutions. Its proprietary technologies cover the full spectrum of automated media intelligence, ranging from text-based entity recognition, relation extraction, topic classification, sentiment analysis, automatic summarization and event detection, to image-based logo and face recognition. Its automatic summarization technology ranked No. 1 in “single document summarization” shared task competition organized by NLPCC’2018. Its topic classification technology finished at the 4th in NLPCC’2018 “Automatic Tagging of Zhihu Questions” shared task. Its technologies have been recognized by top-tier Chinese banks, security firms, media groups and big brands to power their big data enterprise solutions.

SHENYANG YASMART TECHNOLOGY CO. LTD

SHENYANG YASMART TECHNOLOGY CO. LTD. (YASMART) was jointly founded in November 2017 by Northeastern University and NiuTeam, YASMART is devoted to developing natural language processing (NLP) platforms and large-scale data tagging platforms, and provides NLP solutions for vertical industries.
Out of the NLP Lab of Northeastern University, NiuTeam has nearly 40 years’ history of NLP R&D with all the technologies such as word segmentation, POS tagging, NER, information extraction, sentiment analysis, machine reading understanding, and knowledge graph construction, and provides custom NLP services for enterprises.
Currently, YASMART is empowered with large-scale data collcetion (NiuSpider), data tagger (NiuTagger), natural language analysis (NiuParser) and application service (NiuApi). Among them, Niuparser has been acquired by more than 3,000 research institutes worldwide. YASMART has built partnership with the Institute of Infinite Artificial Intelligence (Beijing), Shenyang YaTrans, Tsinghua University, and Beijing DIP, serving the industries of medicine, finance, machine translation and manufacturing.

SoftBank Robotics China: PEPPER & NAO

NAO: Attractive platform, Fully and easily programmable, Versatile, Highly interactive, Multiple sensors, Onboard control, Compatible with standards, Ready to use, Affordable
PEPPER: The First Private Robot which owns Emotion, Feature ① Humanoid Robot which can understand Human's Emotion, Feature ② A Platform that can be Developed, Feature ③ Evolving Cloud Intelligence

Chinese electronic medical record structured analysis system

The electronic medical records are structured using related techniques of natural language processing.
For the unique language style of electronic medical records, the customization of related technologies such as word segmentation, named entity recognition, syntax analysis and dependency analysis is realized.
And through the combination of medical expert knowledge, a medical chart that can express the complete medical process of a patient is constructed from the electronic medical record.
And the automated construction of the medical special disease database has been realized.
It has brought great convenience to the medical research for the information retrieval, screening and queue research.
At the same time, it laid the foundation for clinical decision support based on real cases.

Beijing Alexmasen Technology Co., Ltd.

Beijing Alexmasen Technology Co., Ltd., established in 2012, is a cutting edge high-tech enterprise with Natural Language Processing (NLP), Machine Learning, Knowledge Graph and High-Performance Computing as its core technology advantages.Our company has launched a series of software products, such as "Zhixing Think Tank" and "Gaocan", which is specialized in non-structured data processing. The products are currently applied in the fields of government information inspection, financial regulation, social intelligence collecting, news public opinion and enterprise information monitoring and has acquired dozens of independent intellectual property rights.In line with the spirit of integrating knowledge into practice, the company is committed to transforming scientific research efforts into advanced productive forces and making continuous contributions for the national core scientific and technological development.