Exploring the Human Emotions for Depression Detection from Twitter Data by Reducing Misclassification Rate

A unique sentiment analysis approach is employed for depression detection on Twitter. Leveraging a lexicon ensemble and a novel Neutral Negative Scoring algorithm, user history is analyzed with a neural network, providing personalized recommendations. The Bidirectional Long-Short Term Memory network achieves a remarkable 90% accuracy in detecting depression.

BIoT: Blockchain-based IoT for agriculture

A great attempt on Enhance agricultural transparency with blockchain, consolidating data on events, seed quality, climate, payments, soil moisture, and market prices. Project focuses on storing sensor data in Ethereum blockchain, enabling smart contracts for seamless crop and land transactions.

Recommendation of Crop and Yield Prediction by assessing Soil Health from Ortho-photos

Enhancing farm productivity by recommending crops based on soil nutrient levels and yield predictions. This aids landowners in marketing and storage decisions, benefiting industries collaborating with them. The approach estimates pH and soil nutrients from ortho images to suggest suitable crops for optimal growth conditions.

The Case for Retaining Natural Language Descriptions of Phenotypes in Plant Databases and a Web Application as Proof of Concept

Unlocking shared genetics and stress responses through phenotypic similarities is crucial. Natural language processing offers a resource-efficient alternative to curation-based methods, predicting similar phenotypes and aiding in dataset management. The study introduces phenotype similarity networks and a web app for querying plant gene datasets, showcasing the versatility of these techniques across different situations and species.