Written by Joaquin Melara Commentary from Ezra Duong-Van
KNOWLEDGE IS THE RESULT OF UNDERSTANDING
Dow Jones Factiva news is a powerful business intelligence platform for actionable insights that support strategic decision-making and to proactively identify and respond to opportunities and risk. This intelligence (facts) from news, however, is gleaned by human analysts actually reading to make sense by understanding the content, which remains a time consuming and arduous task.
While Factiva includes powerful search techniques such as ‘codes’ for subject, event, industry, region, language, and company to narrow the qualifying articles for a given purpose, the act of making sense and understanding of the final, albeit reduced set of articles by human readers still remains a bottleneck!
The collaborative effort between Dow Jones Factiva and Lead Semantics squarely targeted this ‘Automated understanding of text’ that results in the knowledge graph extracted from source news. The knowledge graph itself holds the extracted facts for querying and bypasses the actual need for humans’ time to read and comprehend the news for targeted purposes.
“When you deal with unstructured data, you might look at aspects of the document and not the full text to see if you get a better understanding that way. There’s all these little tricks to look at unstructured data to figure out how to get an accurate answer out.
We have a lot of customers who are trying to do this on their own. And they are really struggling with working with unstructured text, how news is reported, and how it affects the business world. That’s still a challenge today. And we’re not familiar with anyone that has the perfect, out of the box solution. There may be one in years to come, but there is none today.
One of the biggest appeals of Lead Semantics is that they enable customers to take a large volume of articles and make sense of them. They offer a very good solution today that does this.”
Ezra Duong-Van on challenges with unstructured data
LEAN APPROACH TO DEVELOPING SUCCESSFUL PILOTS
The ESG domain was chosen as the first focus area for the joint effort recognizing its growing importance as well as recognizing that traditional news monitoring platforms face a challenge in providing insights that integrate ESG factors.
As part of the effort, Lead Semantics’ TextDistil – a modern NLP and Semantic Technology platform was trained to leverage Factiva codes and to ingest the results of Factiva Search to produce a knowledge graph focused on ESG and investment related facts from the news articles.
The resulting graph composed of facts supports two high value use cases; 1) instantly verifiable (augmented) semantic search 2) populating an enterprise Knowledge Graph with facts extracted from news/articles for on-going curation and high value knowledge analytics.
Integration of TextDistil extends Dow Jones Factiva with knowledge graph based analytics and search enabling deeper and comprehensive coverage and analysis of ESG impact of companies and industries.
“Our biggest challenge when encountering customers who are looking to put together projects and use our news to do it is that their success rate is not very high. They do not have the funding, time, and in-house expertise to develop a proprietary solution which has already been solved by TextDistil.
Having Lead Semantics as a partner when we encounter our customers who are going down this journey for a particular domain such as ESG would allow us to have a conversation around their needs. For example, they want to understand the ESG state of thousands of companies such as which are doing better over time, which are doing worse over time, and why.
I know if they were trying to attempt to do that in-house, they would probably do a pilot. They probably would not get meaningful stuff out of it, then they would probably shelf the product. Introducing Lead Semantics into this process would greatly increase the probability of the pilot program being successful.”
Ezra Duong-Van on challenges with pilot programs using news data
“The solution empowers Factiva users to explore voluminous news articles and data through an easy to use TextDistil Search engine which programmatically builds relationships between thousands of concepts mentioned in huge volumes of articles in a matter of minutes and hours rather than days and weeks.”
SETTING OUR SIGHTS ON MODERNIZING NEWS SEARCH
Dow Jones and Lead Semantics embarked on a collaborative journey to develop a potent ESG knowledge graph integration with Factiva. This involved utilizing Lead Semantics’ flagship product, TextDistil.
TextDistil is a modern natural language processing (NLP) and semantic graph technology platform, which is capable of extracting unstructured ESG-related data from news articles and corporate reports. The extracted data is structured into a knowledge graph, where relationships between concepts such as companies, events, and ESG attributes, etc. are computed algorithmically.
ENHANCING FACTIVA WITH AN ESG KNOWLEDGE GRAPH
The integration introduced enhanced search functionalities, allowing users to query for ESG signals in Factiva news using ESG-related keywords, company names, and other specific criteria. The knowledge graph enabled the platform to display interconnected concepts for rich insights, showcasing how news events and ESG considerations intersect.
“Companies’ ESG scores and trends are now visualized in an interconnected browsable way, empowering users to make informed decisions based on timely comprehensive insights.”
“It’s not necessarily just the performance gains, but it’s something that’s repeatable. To have a human do this, they’re a very expensive human. They’re very highly trained. They’re a subject matter expert. They’re very good at doing research. But once I set up a system like this, It can constantly run, and so we can constantly monitor all of these things that are risks or opportunities to the business.
It’s a much more scalable model. And the fact that a human can only read so much, you know, rather than reading 50 articles, we can cover 500 articles or 5000 articles or so on. So It’s being able to expand your ability to not only find a needle in a much bigger haystack, but it’s also to programmatically keep that running on a regular basis at a much lower cost.”
Ezra Duong-Van on Lead Semantics’s flagship product, TextDistil
NEWS WORTHY OUTCOMES
The collaboration resulted in a transformative enhancement of Factiva, aligning it with the evolving landscape of ESG-driven decision-making. The solution empowered users to:
- Monitor and assess companies’ ESG performance in (near) real-time.
- Identify potential risks and opportunities tied to ESG events.
- Understand the correlation between news events and ESG impacts.
- Analyze industry trends and benchmarks in the context of ESG.
- Make data-driven decisions that integrate ESG considerations.
The integration of Lead Semantics’ ESG knowledge graph enriched Factiva’s capabilities, making it a crucial tool for professionals seeking both comprehensive news coverage and insights into ESG factors. Dow Jones’s leadership in business intelligence, responding to market demands and contributing to more informed, sustainable decision-making across industries is bolstered.
The collaboration between Dow Jones and Lead Semantics stands as a testament to the power of innovative partnerships in addressing emerging market challenges. As a result, both companies are able to provide a valuable solution to users and contribute to the advancing the understanding and integration of ESG considerations in the business landscape.
“Lead Semantics’s flagship product, TextDistil can be used for a lot of different things… It is not designed for one industry or one topic. By taking the news, we can distill it down to an industry or a topic and then Lead Semantics can then very quickly go through all of those articles and understand what the trends are that we’re looking for. Anything that’s captured in the news is a good candidate for this, and the use cases are all over the map”.
Ezra Duong-Van on TextDistil’s versatility
“Anything that’s captured in the news is a good candidate… the use cases are all over the map”
