Updates in July. 19

You can download a baseline code from the organizer here or on the data webpage.

Background

Information extraction is the task of extracting specific events or facts from natural language texts. It helps us automatically classify, extract and reconstruct massive content. Intelligent text extraction is an important basis for artificial intelligence applications such as information retrieval, question answering, and dialogue. It can overcome difficulties of understanding the unstructured and ambiguous text, and mine and capture the valuable information from it. The information can then be used for business consultation, decision support, precision marketing, etc., therefore practically significant to the industry.

DataGrand is China’s leading intelligent text processing enterprise. Using the most advanced semantic analysis technology, DataGrand provides A.I. engines for automatic text extraction, text review, error correction, search, recommendation, writing, and others, allowing computers to automate a wide range of tasks normally handled by humans and maximize business efficiency.

DataGrand' s core team consists of senior executives and technical experts from illustrious companies such as Tencent, Shanda, Baidu, and AliBaba. Together, DataGrand has won first prizes in multiple ACM international data mining competitions, applied for more than 60 patents, and published two books and hundreds of technical papers on artificial intelligence. DataGrand is also a proud member of Microsoft Accelerator, Legend Star, and SAP Innovation Camp, and maintains joint research with the text analytics laboratory at Fudan University. DataGrand’s market accomplishments are equally remarkable; having received hundreds of millions in investments from distinguished institutions such as Zhen Fund, SAIFPartners, and F&G Venture, DataGrand is China’s most highly-invested text analytics enterprise. DataGrand has also successfully provided text analytics services to many well-known enterprises such as Huawei, Haier, Changhong, Jingdong, China Merchants Bank, and Huatai Securities.

Task

The task of this competition is to extract 3 different fields based on labeled text and large-scale unlabeled corpus. The organizers hope all participants use state-of-art NLP and AI technologies to deeply mine the features of words and sentences, construct language models and precisely extract information.