Financial　theory　holds　that　attention　plays　a　significant　role　in　the　information　response,　and　internet　attention　has　been　used　widely　to　explore　their　influence　on　stock　market　microstructure.　Based　on　the　thermal　optimal　path　method　and　social　network　data,　this　paper　constructs　two　dynamic　variables　including　investor　and　media　attention　reflecting　the　internet　attention　and　examines　the　lead-lag　relationship　between　the　internet　attention　and　market　liquidity　measures.　Furthermore,　the　sample　is　further　divided　into　three　parts　including　stationary　and　fluctuation　periods　to　explore　the　predictive　ability　of　internet　attention　effects　on　Chinese　stock　market　liquidity.　The　main　results　reveal　that　the　lead-lag　orders　between　the　internet　attention　and　market　liquidity　are　not　always　positive　or　negative.　In　other　words,　the　internet　attention　does　not　always　dominate　the　stock　market　liquidity,　and　vice　versa.　Moreover,　there　are　significant　differences　in　the　results　of　lead-lag　orders　between　three　different　subsample　periods.　The　empirical　results　confirm　that　internet　attention　facilitates　forecasting　market　performance　in　the　Chinese　stock　market,　and　supplements　the　relevant　theories　of　stock　market　trading　and　behavioral　finance　from　the　perspective　of　econophysics.　(C)　2020　Elsevier　B.V.　All　rights　reserved.