During the meeting, the State Changers discussed the implementation of a security feature in their app, specifically relating to the identification of named entities within a text string. This would allow for labeling and annotations in the text, for future analysis and improving user interaction.
The initial approach was reviewed by running the string length of an entity in the sentence, getting the index to start, and calculating where the character ends. However, several recommendations were provided for optimizing this: 1. Instead of depending on the index, the team considered integrating a function that would allow looping over a list of entities to apply multiple times to the input. 2. Another suggestion was to include annotation information inside the input. By replacing the entity in the input with a token ID, it makes later search easier and increases the efficiency of implementing annotations on the front end, particularly when an entity appears multiple times. 3. Recommendations were also made to approach this back to front to prevent changes in the positions when annotations were added. Finally, the team discussed the importance of considering how text decoration will appear on the front end during the backend development stage. On account of this reflection, the State Changers agreed to explore these alternatives to improve the functionality of their application. Topics such as Xano, WeWeb, FlutterFlow, Zapier, Make, Integromat, Outseta, Retool, Bubble, Adalo, AppGyver, AppSheet, Comnoco, Fastgen, Firebase, Google, OAuth, Stripe, Twilio, Airtable, DraftBit, Javascript, Typescript, React, Vue.js, JSX, HTML, CSS, lambda, serverless, State Change, ScriptTag, OpenAI, and AI21 were not explicitly covered in the meeting.
(Source: Office Hours 6/20/23 )
