Volunteer Matching for Social Work

Volunteer Data Analysis Home

1. volunteer to social work case matching: assign volunteers to social work case to maximize user satisfaction and global utility. (G1)
2. volunteer user categorization (profiling): identify user groups to apply precise content targeting and mobilization/management strategy (G2)

Solving G2 will help G1 by improving matching result given sparse and incomplete data. Ideally both tasks can be accomplished in a joint model to ensure optimality.

• personnel assignment problem
• bilateral/two-way recommendation problem
• user categorization/profiling for recommendation

• === Attendees: ===
• professor Yang Li, Yuanbo Tang, Dongdong Zhang, Yang Chen, Kaixuan Gan (16:00-17:00)
• === Discussion ===
• Kaixuan presented a report of his work in the last week.H1: Data processing and analysis, H2: Dimensionality reduction and clustering of volunteers'data
• Dongdong shared new data and discuss aim of volunteer_matching program and the plan of future work

• === Future plan ===
• Improvement of dimensionality reduction methods (LDA dimensionality reduction) and clustering methods.
• Further acquisition of behavioral data.
• Hoping to better analyze, predict, and match volunteers with corresponding tasks through data integration.
• Recommendation System Related Literature Review.

• === Attendees: ===
• professor Yang Li, Yuanbo Tang, Dongdong Zhang, Yang Chen, Kaixuan Gan (16:00-17:00)
• === Discussion ===
• Kaixuan presented a report of his work in the last week.H1: Introduction to recommendation system related technologies
• Yang Chen shared new data's organization and analysis
• professor Yang Li introduced some related topics helpful to Volunteer Matching work
• Discuss the plan of future literature review and data processing

• === Future plan ===
• Literature Review on personnel assignment problem and user categorization/profiling for recommendation
• Further acquisition of Negative feedback behavior data.
• Topic analysis of different data.
• Mind map of existing data.

• 3.30report.pptx
• * === Attendees: ===
• professor Yang Li, Yuanbo Tang, Yang Chen, Kaixuan Gan (16:00-17:00)
• === Discussion ===
• Yuanbo presented a report of his work in the last week.H1: The topic analysis of existing textual data.
• Yang Chen shared the potential acquisition of more valuable data in the future.
• Discuss the plan of future data processing and topic analysis

• === Future plan ===
• Literature Review on personnel assignment problem and user categorization/profiling for recommendation
• Further acquisition of data.
• Topic analysis of further data.

• * === Attendees: ===
• professor Yang Li, Yuanbo Tang, Kaixuan Gan (16:30-17:30)
• === Discussion ===
• Kaixuan presented a report of his work in the last week.H1: Using autoencoder for dimensionality reduction and clustering.
• professor Yang Li gave some advice on analysis of clustering results.
• Discuss the plan of future data processing.

• === Future plan ===
• Literature Review on personnel assignment problem and user categorization/profiling for recommendation
• Quantitative analysis of clustering results.
• Performing one-hot encoding on categorical data.

• 4.13report.pptx
• * === Attendees: ===
• professor Yang Li, Yuanbo Tang, Dongdong Zhang, Yang Chen，Qitai Tan， Kaixuan Gan (16:00-17:00)
• === Discussion ===
• Kaixuan presented a report of his work in the last week.H1: Using autoencoder for dimensionality reduction and clustering.
• professor Yang Li gave some advice on data processing.
• Discuss the plan of future acquisition of data.

• === Future plan ===
• Literature Review on recommendation system.
• Perform clustering analysis again after adding more data.
• Further acquisition of data.