AI Content Chat (Beta) logo

Preqin ESG Solutions 57 Step 3: If a private company or infrastructure asset is not identi昀椀ed as a “Very Likely” source of impact in the 昀椀rst two steps, its impact Leverage potential is estimated through our NLP model. natural Preqin leverages its database of company and infrastructure language asset descriptions to form a view of a company’s activities that goes beyond raw industry classi昀椀cation. These asset processing descriptions tell us about a company’s main area of operations, its supply chains and how it conducts business. They also (NLP) to reveal information about shareholders’ values, aspirations, and objectives. Excepting direct experience as a client, contractor identify or employee, these asset descriptions are the best source of information about the nature and characteristics of a given potential enterprise. impact Our method for assigning likelihood levels of impact potential assets to assets neither operating in known impact industries nor featuring in impact funds is outlined below: a. Using the corpus originating from the UN SDG’s website and o昀케cial documentation we generate a list of frequently used keywords. This list is then curated to remove words unrelated to impact (e.g. “company”), and add words that are missing (such as “sustainability”). The objective is to create an initial list of keywords associated with impact. b. We then stem each keyword to account for di昀昀erent grammatical forms and variants of each word (so “empowerment” would be replaced with “empower*” – the asterisk representing any sequence of characters). c. We then rate each stemmed keyword, based on its frequency throughout the SDGs objectives, normalizing these scores to percentages between zero and hundred. d. For each private company on Preqin Pro: i. We search its description for words matching the keywords from point (b) ii. We then add the scores (from point c) of each stemmed keyword that appears in the description. Each stemmed keyword is counted only once, irrespective of the number of repetitions ↗ Back to Contents

Preqin ESG Solutions Methodology - Page 57 Preqin ESG Solutions Methodology Page 56 Page 58