Big　consumer　opinion　data　provide　valuable　information　about　customer　preferences.　These　online　opinions　facilitate　designers　to　capture　customer　requirements　(CRs)　and　understand　customer　satisfaction　(CS)　for　new　product　development　(NPD).　A　visible　gap　between　practical　significance　and　research　studies　is　to　uncover　nonlinear　relations　between　CRs　and　CS　and　derive　strategic　suggestions.　Accordingly,　a　framework　for　online　CRs　Kano　categorization　is　proposed.　Firstly,　both　explicit　and　implicit　features　are　extracted　from　opinionate　texts　to　better　capture　CRs.　Secondly,　to　evaluate　the　impact　of　CRs　on　the　overall　CS,　a　multi-layer　neural　network　is　invited,　in　which　the　impact　from　both　positive　and　negative　opinions　over　each　product　feature　are　distinguished.　Finally,　according　to　the　estimated　impact,　CRs　are　categorized　by　a　Kano　Model　based　approach.　To　evaluate　the　effectiveness　of　the　proposed　framework,　a　case　study　that　analyzes　a　large　number　of　phone　reviews　is　presented.　Categories　of　studies　were　benchmarked　to　demonstrate　the　competitiveness　of　utilized　approaches.　This　study　is　argued　to　disclose　complex　relations　between　CRs　and　the　overall　CS　as　well　as　strategic　improvement　suggestions　by　online　opinion　analysis.　It　enlightens　designers　to　infer　constructive　strategies　from　big　consumer　opinion　data　for　market-driven　NPD.