[Purpose/　significance]　A　multi-　task　user　needs　identification　model　integrating　disease　characteristics　(MUNI-DC)　is　proposed　for　user　generated　content　in　online　health　communities.　Deeply　explore　user　needs　and　form　a　demand　theme　system　consisting　of　two　parts:　questioning　intention　and　questioning　entity.　[Method/　process]　By　constructing　a　BERT-wwm　model　for　pre-training　user　demand　data　and　integrating　user　medical　condition　characteristics　to　achieve　recognition　of　user　intentions;　Furthermore,　a　multi-layer　label　pointer　network　is　used　to　achieve　entity　recognition　of　user　query　data　in　online　health　communities.　Based　on　this,　identify　the　needs　of　users　in　online　health　communities.　[Result/　conclusion]　Comparative　experiments　have　shown　that　compared　to　a　single　task　model,　this　model　has　improved　in　precision,　recall,　F1,　and　other　indicators　of　user　requirement　recognition　results.　The　ablation　experiment　found　that　the　fusion　of　disease　characteristics　and　multi-layer　label　pointer　network　can　effectively　improve　the　user　demand　recognition　performance　of　the　model.　The　proposed　MUNI-DC　model　has　reference　value　in　dealing　with　online　health　community　user　information　demand　analysis　tasks.　©　2025　Information　studies:　Theory　and　Application.　All　rights　reserved.