金融学中的知识图与自然语言Knowledge Graphs & Natural Language in Finance |
|
课程网址: | http://videolectures.net/www2021_kambadur_knowledge_graphs/ |
主讲教师: | Anju Kambadur |
开课单位: | 彭博公司 |
开课时间: | 2021-05-03 |
课程语种: | 英语 |
中文简介: | 金融决策涉及发现和总结相关信息、产生交易想法、寻找流动性和交易对手、进行事后分析以及发布报告。在彭博社,我们利用机器学习、知识图表和语言技术的最新发展,使我们的客户能够以智能方式在其决策过程的每一个步骤中以高精度、低延迟、大规模地获得市场优势。在本次演讲中,我们将介绍彭博知识图和高级自然语言处理(NLP)技术在以下领域的使用:(a)信息和关系提取,以帮助我们的记者自动生成新闻,(b)命名实体识别和链接、主题分类、聚类和摘要,以帮助我们的客户消费内容,(c)语言建模和语义分析,以促进基于自然语言的内容发现;(d)对话理解,以帮助构建即时消息(场外交易的主要媒介),以及(e)情绪分析,以生成用于阿尔法生成的结构化时间序列信号。在整个演讲中,我们将重点介绍我们小组发表的文章,这些文章可以作为进一步的阅读材料。 |
课程简介: | Decision-making in finance involves discovering and summarizing relevant information, generating trade ideas, finding liquidity and counterparties, performing post-hoc analyses, and publishing reports. At Bloomberg, we leverage recent developments in machine learning, knowledge graphs, and language technology to enable intelligent ways for our clients to obtain market advantage in every step of their decision making process, at scale, and with high precision and low latency. In this talk, we cover the use of the Bloomberg Knowledge Graph and advanced natural language processing (NLP) techniques in the following areas: (a) information and relationship extraction to assist our journalists with automated news generation, (b) named entity recognition and linking, topic classification, clustering, and summarization to assist our clients in consuming content, (c) language modeling and semantic parsing to facilitate natural-language-based discovery of content, (d) dialog understanding to aid in structuring instant messages, the primary medium for over-the-counter trading, and (e) sentiment analysis to generate structured time-series signals for alpha generation. Throughout the talk, we will highlight articles published by our group that can serve as further reading material. |
关 键 词: | 金融; 知识图表; 语言技术 |
课程来源: | 视频讲座网 |
数据采集: | 2022-04-13:zkj |
最后编审: | 2022-04-13:zkj |
阅读次数: | 27 |