Legal teams dealing with E-discovery face a range of problems:
Mass volumes of documents of which only a small portion are relevant.
Processing of the documents includes extraction of entities, ideas, affinities between them, categorization and prioritization.
Strict time frames in which documents need to be submitted for the court’s scrutiny mean that it is crucial for the process to be able to expeditiously, with minimal expenses of manual review.
The need for coordination within a team means that the system must link the dots to find the relevant correspondence on one hand, and screen out duplicates to reduce time and manpower, on the other hand.
E-discovery of documents today is geared primarily to extract “key words” (string search) finding specific documents that the searcher knows in advance that are needed. However, the proverbial “unknown unknown” is no less – and frequently even more – important for a case.
The IntuView Solution
IntuView’s meaning mining tool – IntuScan can save legal practice huge amounts of time and labour costs. It enables the user to identify salient information under time constraints from massive amounts of data with minimal reliance on human triage. While data mining simply extracts key words from big data, meaning mining derives their significance.
IntuScan ingests large amounts of documents, to extract from the documents all entities, relationships between entities, sentiments, events and other salient information, categorizes and prioritizes the documents and feeds the extracted information into a semantic database and into reports that succinctly summarize the information of the texts. The user may then query the database for texts that contain any combination of information – even if the user does not know exactly what he or she is looking for.
IntuScan quickly analyses large amounts of documentation in a variety of supported languages, including: English, French, Spanish, Arabic, Urdu, Russian, and Indonesian, with Hindi and Farsi on the drawing board. It uses intuitive artificial intelligence to replicate the methods by which the human brain not only recognises relevant words and phrases, but also places them within a cultural, historical and semantic context to figure out the meaning implied by the creator of the document.
The IntuScan capability is particularly useful for the disclosure process of litigation cases, where huge volumes of text are exchanged. IntuScan allows the legal team to identify all the entities – persons, organisations, addresses, bank accounts, email addresses and their owners, places, etc. – mentioned in all the documents, without the need to input in advance lists of keywords or names. At the same time, IntuScan performs automated classification of the documents to issues, areas of competence of the different members of the team, priority , nd relevance. These categories are pre-defined at the beginning of the process and automatically updated in a learning process in the light of the course of the discovery and decisions made by the users. These capabilities allow the legal team to peruse the information in the discovery documents in order to identify likely directions of investigation at an early stage of the process.
IntuScan does not replace human experience and intelligence, but rather serves as an expert system that reduces the number of irrelevant and superfluous documents that the legal team has to peruse and allows its members to focus on the key information.
One of the consequences of globalization is the emergence of corpora of regulations issued by national and trans-national regulatory bodies (the EU, International treaties etc.). These regulations are updated regularly and do not lend themselves to easy interpretation. Hence, commercial or financial institutions or companies may easily find themselves in non-compliance of new regulations that apply to them. The problem therefore is to find, within the enormous corpus of regulations that are issued, which ones apply to which department in a commercial body. The domains that this problem applies to are manifold: from banking, insurance, transportation and shipping, and more.
IntuView has applied its technology to this task in order to simplify the search in regulations and to automatically flag those that may apply to certain users under certain conditions. To this end, IntuView builds a domain-specific ontology for the domain of the regulation, analyses the texts of the regulations and inserts them into a database. The different divisions of a client’s company define their areas of competency and responsibility and these serve as automated queries to the database and return the articles of the regulations that apply to those divisions.
IntuView is also collaborating with experts of maritime regulations in a joint venture to apply this methodology to maritime regulations to improve the safety and efficiency of ships arriving in ports and their compliance with safety regulations.