The Hype Cycle

In 2018, knowledge graphs (KGs)was on the rise within the hype cycle alongside AI and few other emerging technologies. Knowledge graph is here to fix the problem of information overload, driving the information economy to knowledge economy.

What are the main use cases of KGs in investment?
  1. QA systems

  2. Semantic search for research and knowledge discovery

  3. Dynamic risk analysis

  4. Content-based recommendation engines

  5. Knowledge arbitrage

  6. Thematic investing

  7. Knowledge management systems

  8. Enterprise data governance

  9. Risk exposure

Describe enterprise data governance.

KGs can be use to consolidate and centralise all the isolated data sources by processing different types of sources and extracting units of knowledge. KGs can project information into a multidimensional conceptual space where we can use similarity measures along different dimensions to group related concepts.

Describe risk exposure.

Complex contagion is the concept that multiple sources of exposure are required for an individual to change his/her behaviour. Measuring risk exposure is all about complex contagion. There are many direct and indirect factors that affect a company, industry, or economy. KGs can be used for reasoning and inferring relationships between entities.

Yewno’s AI Financial platform allows you to quantify your portfolio exposure to complex concepts, tracking the relationships between entities in your portfolio and the concept / event.

What’s the argument for knowledge graph as an alternative data engine?

Information is not the same as knowledge. And knowledge does not automatically translate into actions. Many hedge funds and asset management firms indicate that not only do they need to rely on traditional data such as fundamentals and pricing data but also on non-traditional data about People, Governance, Events, and Transcripts. All these datas have their own format and KGs is the perfect tool to consolidate all of them together.

What are some of the key challenges in AI-based investment and how can KGs help solve them?

There are three main issues:

  1. Fairness

  2. Accountability

  3. Transparency

Knowledge graph encodes the meaning of data and can be self-descriptive, meaning that KGs can offer transparency and interpretability that could lead to fairness and accountability.



Data Scientist

Leave a Reply