Online　question　and　answer　(Q&A)　community　users＇　knowledge　contribution　behaviour　was　studied　using　primary　and　secondary　data　and　different　research　approaches.　However,　this　topic　is　never　explored　in　the　context　of　content　(knowledge)　shared　in　these　communities.　Furthermore,　online　social　interaction＇s　role　as　a　mediator　is　also　ignored　in　online　Q&A　communities.　This　study　model　explored　community　recognition,　online　social　interaction,　devotion　to　community,　self-satisfaction,　and　a　sense　of　reciprocation＇s　role　in　the　knowledge　contribution　behaviour　of　Q&A　community　users.　We　collected　709　online　Q&A　community　users＇　responses　and　used　SEM-ANN　two-stage　hybrid　approach　to　capture　linear　and　nonlinear　relationships　between　variables.　Results　revealed　that　all　explanatory　variables　are　positively　significant,　while　the　sense　of　reciprocation　is　negatively　significant　to　knowledge　contribution.　It　strengthens　the　earlier　researcher＇s　claim　that　the　term　＇tragedy　of　common＇　implies　online　Q&A　communities.　Normalised　importance　results　in　the　second　stage　figuring　out　that　community　recognition,　online　social　interaction,　and　community　devotion　are　the　most　influential　factors　behind　knowledge　contribution　in　online　Q&A　communities.　Findings　amplify　our　apprehension　about　the　knowledge　contribution　behaviour　of　Q&A　community　users.　It　also　provides　evidence　that　dual-stage　deep　learning　modelling　can　better　capture　variables＇　linear　and　nonlinear　relationships.