Document parsing and text mining in Python

Programmer for Tree Parsing/Text Mining Job Summary Seeking an experienced programmer for engagement in long-term freelance work. Strong tree parsing skills are essential. A background in NLP and experience with NLTK is preferred but not required. Pay is commensurate with experience and is hourly-based. As part of our hiring process, we ask that interested candidates successfully complete the tasks below to demonstrate basic competency. Project Background The SEC stores various text files they receive from companies on their Edgar website. The files typically contain detailed discussions of companies’ performance as well as financial data summarizing their performance. Attached is a random sample of 15 full .txt files from 5 different years with a file type of “10-K” from Edgar. You will find files which embed HTML, SGML, or XBRL code, in addition to tables, special characters, images, and other embedded files, such as PDF, etc. Tasks Extract the following sections from the 10-K using a tree parser: Management Discussion and Analysis (MD&A), Risk Factors, and Notes to the Financial Statements. Flatten each section extracted to raw text. That is, remove all code, tables, images, or embedded files. Write the raw text of each section to a separate .txt file. The filename for the raw text file should be that of its parent with a suffix for each section appended (e.g., “*[login to view URL]”, “*[login to view URL]”, and “*[login to view URL]”). Discuss any outstanding issues, questions, or concerns regarding the steps above. For example, discuss weaknesses in your approach to identifying section and sentence boundaries. Apply For full consideration, please upload your resume, output, and responses by April 15, 2015. We are an equal opportunity employer. Work permits or visas are not required.