To put it simply, semantic search stands to revolutionize financial markets.
Everyone today knows that they can easily find a list of relevant information and potential answers to any question on their mind by simply typing a thread of words into their internet search engine of choice. For financial professionals, this has not historically been the case as finding data has relied on using arcane codes and pulling together information from disparate pages. But that is all changing as the financial markets -- an industry that is all about data -- begins to realize the opportunity that search offers. In an industry bombarded with data, the advent of semantic search in financial markets stands to force a shift in the way financial professionals consume and analyze information and, most importantly, make money.
From data to answers
The consumer search experience has gone through significant evolution in the past few years with the most innovative and successful consumer search products using natural language techniques and artificial intelligence, include machine learning, to move from the information space to the answer space.
However search technology in the financial markets industry is really still in its infancy. Restricted by legacy technologies going back 20 or even 30 years, search features have been basic, with a heavy reliance on proprietary keywords and syntax rather than natural language. There is therefore a steep learning curve for new users and discovery of information is impaired because information helpful to a user's workflow potentially goes undiscovered without the right navigation path or shortcut code.
[For more on market data platforms are making better use of natural language search, read: Semantic Search Takes Thomson Reuters Eikon to New Levels].
Users in the financial industry are obviously also consumers of the public internet, and over the past few years they have noted the discrepancy in their information retrieval experiences. They are seeing advances like autosuggest and keyword news search, however their experience in the work environment has been that information lookup and idea generation is still a tedious task.
Time For Semantics
Therefore, it is easy to conclude that the time is ripe to revolutionize financial markets by taking advantage of the advances being made with internet search. Given the bounded vocabulary and rich metadata prevalent in financial data, this is where semantic search techniques really come into their own.
Semantic search technology steps beyond keyword search to improve the accuracy of search results -- it uses the contextual meaning of the search terms to understand the searcher's intent. For example, semantic search can enable a product to return the correct news articles when terms like 'earnings,' 'revenues,' or 'rebase' are recognized. It can help financial-focused queries answer questions such as 'pepsi coke spreads.'
Natural language processing can additionally be used to properly understand questions as expressed naturally by the user. A system should understand parts of speech, and be able to recognize words like 'what,' 'when,' and 'where' to determine the type of question. It can also pick up on words like 'for,' 'in,' or 'between' to properly connect various parts of the search phrase into something meaningful.
The above two capabilities enable the product to achieve search nirvana -- providing exact answers to questions.
So the natural evolution of search means that when a user requests a revenue number, we are now looking to provide a single number, rather than point users to the income statement. No longer should users be taken to pages where the answer may be buried among many other facts. No longer should they be faced with multiple results to click through until they find the right one.
Search is a learned behavior and confidence builds with success. With the introduction of a question answering system, a user's confidence can grow and many new scenarios will be possible. Where previously they may have felt daunted by tasks that required multiple steps, they will now feel emboldened to ask a simple question.
With search queries reduced to seconds, the paradigm shift occurs and a new generation of financial products will be born.
The Next Paradigm Shift
The goal with search is to quickly get users to the information they seek, shortening their path to insight. Users need to be able to find everything right from a single search box, using natural language, rather than having to learn a specific syntax.
Not only does this offer a massive time saving, but it has a binary impact. If you are an investment manager, analyst or salesperson looking for trade ideas and you have a query that would take hours to explore but you're 75% sure it's a dead end, you probably won't ask it. If it takes you four seconds to explore the idea, you might as well make the query. As semantic search introduces financial markets professionals to a world of exact answers, we will witness expanded exploration of trading ideas. With one click, financial professionals can weed through all of the noise in market data. This is about throwing away the time-consuming user manuals and empowering the markets to do what they do best: make money.
About The Author:
Haris Husain is Head of Search for the Financial & Risk business of Thomson Reuters. He is responsible for the strategy and development of search capabilities within Thomson Reuters Eikon desktop and mobile products. As head of search, he leads the firm's efforts to develop intelligent search tools that deliver an exact answer to natural language queries, navigating customers through the breadth and depth of data available in Eikon.