| <!DOCTYPE html> |
| <html> |
| <head> |
| <title>DeReKo-Word-Vector-Distances: <%= $word %></title> |
| <link rel="stylesheet" href="//code.jquery.com/ui/1.12.1/themes/base/jquery-ui.css"> |
| <link href="https://fonts.googleapis.com/css?family=Lato|Roboto+Condensed" rel="stylesheet"> |
| <script src="http://code.jquery.com/jquery-latest.min.js"></script> |
| <script src = "https://cdn.datatables.net/1.10.16/js/jquery.dataTables.min.js"></script> |
| <script src = "https://cdn.datatables.net/fixedcolumns/3.2.3/js/dataTables.fixedColumns.min.js"></script> |
| <script src = "https://cdn.datatables.net/plug-ins/1.10.16/sorting/scientific.js"></script> |
| <script src='https://cdnjs.cloudflare.com/ajax/libs/mathjax/2.7.2/MathJax.js?config=TeX-MML-AM_CHTML'></script> |
| <link rel="stylesheet" href="https://cdn.datatables.net/1.10.16/css/jquery.dataTables.min.css"> |
| <link rel="stylesheet" href="/derekovecs/css/derekovecs.css"> |
| <script |
| src="http://code.jquery.com/ui/1.12.1/jquery-ui.min.js" |
| integrity="sha256-VazP97ZCwtekAsvgPBSUwPFKdrwD3unUfSGVYrahUqU=" |
| crossorigin="anonymous"></script> |
| <script> |
| MathJax.Hub.Config({ |
| config: ["MMLorHTML.js"], |
| jax: ["input/TeX","input/MathML","output/HTML-CSS","output/NativeMML", "output/PreviewHTML"], |
| extensions: ["tex2jax.js","mml2jax.js","MathMenu.js","MathZoom.js", "fast-preview.js", "AssistiveMML.js", "a11y/accessibility-menu.js"], |
| TeX: { |
| extensions: ["AMSmath.js","AMSsymbols.js","noErrors.js","noUndefined.js"] |
| } |
| }); |
| var urlParams = new URLSearchParams(window.location.search); |
| var currentWords = urlParams.get("word"); |
| var CIIsearchWords = (currentWords && currentWords.includes(" ") ? '('+currentWords.replace(/ +/g, " oder ")+')' : currentWords); |
| var collocatorTable = null; |
| |
| $(document).ready(function() { |
| |
| $('#firstable').hide(); |
| //Set up a callback to hear back when MathJax is done rendering the equations |
| // it finds |
| $('#ccd').load( |
| '@Url.Action("ActionResultMethod","ControllerName",{controller parameters})', |
| function () { |
| MathJax.Hub.Queue( |
| ["Typeset",MathJax.Hub,"ccd"], |
| function () { |
| $("#lfmd_tt").attr("title",$("#lfmd_ttt").html()); |
| $("#npmi_tt").attr("title",$("#npmi_ttt").html()); |
| } |
| ); |
| }); |
| |
| //set things up so that we can shove raw html into what is shown in the tooltip; |
| // in this case, we will have already put into the title attribute the html that |
| // contains the MathJax rendered equations (via what we do in the callback). |
| $(function () { |
| $(document).tooltip({ |
| content: function () { |
| return $(this).prop('title'); |
| } |
| }); |
| }); |
| |
| $("input").bind("keydown", function(event) { |
| // track enter key |
| var keycode = (event.keyCode ? event.keyCode : (event.which ? event.which : event.charCode)); |
| if (keycode == 13) { // keycode for enter key |
| // force the 'Enter Key' to implicitly click the Update button |
| document.getElementById('SEARCH').click(); |
| return false; |
| } else { |
| return true; |
| }}); |
| |
| var collocatorTable_activated = false; |
| $( "#tabs" ).on( "tabsactivate", function( event, ui ) { |
| if (localStorage) localStorage['tab'] = ui.newTab.index(); |
| if(ui.newTab.index() == 2 && !collocatorTable_activated) { |
| classicCollocatorTable.columns.adjust(); |
| collocatorTable.columns.adjust(); |
| collocatorTable_activated = true; |
| } |
| }); |
| |
| $(function(){ |
| $("#SEARCH").click(function() { |
| window.open($(location).attr('pathname')+'?'+$('form').serialize(), "_self"); |
| }); |
| }); |
| |
| function changeCharColor(txt, heat, word) { |
| var newText = ""; |
| for (var i=0, l=txt.length; i<l; i++) { |
| newText += (i == 5 ? txt.charAt(i) : '<a href="http://korap.ids-mannheim.de/kalamar/?ql=cosmas2&q=' + |
| CIIsearchWords + ' /' + (i > 5? '%2B' : '-') + 'w' + |
| Math.abs(i-5) + ':' + Math.abs(i-5) + ' ' + word + |
| '" target="korap"><span style="background-color:' + |
| getHeatColor(heat[i]/maxHeat)+'">'+txt.charAt(i)+'</span></a>'); |
| } |
| return newText; |
| } |
| |
| function getHeatColor(value) { |
| var hue=((1-value)*120).toString(10); |
| return ["hsl(",hue,",90%,70%)"].join(""); |
| } |
| |
| function bitvec2window(n, heat, word) { |
| var str = n.toString(2).padStart(10, "0") |
| .replace(/^([0-9]{5})/, '$1x') |
| .replace(/0/g, '·') |
| .replace(/1/g, '+'); |
| return changeCharColor(str, heat, word); |
| } |
| |
| % use Mojo::ByteStream 'b'; |
| var paraResults = <%= b(Mojo::JSON::to_json($lists)) %>; |
| var urlprefix = new URLSearchParams(window.location.search); |
| if (paraResults.length > 0 && paraResults[0] != null) { |
| var nvecs = [], |
| nwords = [], |
| nranks = [], |
| nmarked = []; |
| for(var i = 0; i < paraResults.length; i++) { |
| nwords = nwords.concat(paraResults[i].map(function(a){return a.word;})); |
| nvecs = nvecs.concat(paraResults[i].map(function(a){return a.vector;})); |
| nranks = nranks.concat(paraResults[i].map(function(a){return a.rank;})); |
| nmarked = nmarked.concat(paraResults[i].map(function(a){return a.marked;})); |
| } |
| showMap({target: " "+urlParams.get('word')+" ", mergedEnd: 0, words: nwords, vecs: nvecs, ranks: nranks, marked: nmarked} ); |
| var t = $('#firsttable').DataTable({ |
| data: paraResults[0], |
| "sScrollY": "780px", |
| "bScrollCollapse": true, |
| "bPaginate": false, |
| "bJQueryUI": true, |
| "dom": '<"top">rt<"bottom"flp><"clear">', |
| "initComplete":function(settings, json) { |
| $('td.paradigmator a').on('mousedown', function(e) { |
| return paradigmatorClick(e, paraResults[0][0].word, this.childNodes["0"].textContent); |
| }); |
| }, |
| "columns": [ |
| { "data": "rank", type: "allnumeric" }, |
| { "data": "dist", render: function ( data, type, row ) {return data.toFixed(3) }}, |
| { "data": "word", class: "paradigmator", render: function ( data, type, row ) {urlprefix.set("word", data); return '<a href="?' + urlprefix + '">' + data + '</a>' }} |
| ], |
| "columnDefs": [ |
| { className: "dt-right", "targets": [0,1] }, |
| { "searchable": false, |
| "orderable": false, |
| "targets": 0 |
| }, |
| { "orderSequence": [ "desc" ], "targets": [ 1 ] }, |
| { "orderSequence": [ "asc", "desc" ], "targets": [ 2 ] }, |
| ], |
| "oLanguage": { |
| "sSearch": "Filter: " |
| }, |
| "order": [[ 1, 'desc' ]], |
| } ); |
| |
| t.on( 'order.dt search.dt', function () { |
| t.column(0, {order:'applied'}).nodes().each( function (cell, i) { |
| cell.innerHTML = i+1; |
| } ); |
| } ).draw(); |
| |
| $( "#first" ).clone().prependTo( "#tabs-2" ); |
| |
| } |
| |
| |
| var collocatorData = <%= b(Mojo::JSON::to_json($collocators)) %>; |
| var maxHeat; // = Math.max.apply(Math,collocatorData.map(function(o){return o.cprob;})) |
| |
| if (collocatorData != null) { |
| maxHeat = Math.max.apply(Math,collocatorData.map(function(o){return Math.max.apply(Math,o.heat);})) |
| collocatorTable = $('#secondtable').DataTable({ |
| data: collocatorData, |
| "sScrollY": "780px", |
| "bScrollCollapse": true, |
| "bPaginate": false, |
| "bJQueryUI": true, |
| "dom": '<"top">rt<"bottom"flp><"clear">', |
| "columns": [ |
| { "data": "rank", type: "allnumeric" }, |
| { "data": "pos", width: "7%", sClass: "dt-center mono compact", render: function ( data, type, row ) {return bitvec2window(data, row.heat, row.word) }}, |
| { "data": "max", render: function ( data, type, row ) {return data.toFixed(3) }}, |
| { "data": "average", render: function ( data, type, row ) {return data.toFixed(3) }}, |
| { "data": "prob", type: "scientific", render: function ( data, type, row ) {return data.toExponential(3) } }, |
| { "data": "cprob", type: "scientific", render: function ( data, type, row ) {return data.toExponential(3) } }, |
| { "data": "overall", type: "scientific", render: function ( data, type, row ) {return data.toExponential(3) } }, |
| { "data": "word", sClass: "collocator" }, |
| { "data": "rank", type: "allnumeric" } |
| ], |
| "columnDefs": [ |
| { className: "dt-right", "targets": [0,2,3,4,5,6] }, |
| { className: "dt-center", "targets": [ 1] }, |
| { "searchable": false, |
| "orderable": false, |
| "targets": [0, 8] |
| }, |
| { "type": "scientific", targets: [2,3,4,5,6] }, |
| { "orderSequence": [ "desc" ], "targets": [ 2, 3, 4, 5, 6 ] }, |
| { "orderSequence": [ "asc", "desc" ], "targets": [ 1, 7 ] }, |
| { "targets": [8], "visible": false } |
| ], |
| "oLanguage": { |
| "sSearch": "Filter: " |
| }, |
| "order": [[ 4, 'desc' ]], |
| } ); |
| $.ajaxSetup({ |
| type: 'POST', |
| timeout: 30000, |
| error: function(xhr) { |
| $('#display_error') |
| .html('Error: ' + xhr.status + ' ' + xhr.statusText); |
| } |
| }); |
| |
| |
| if($('#sprofiles').length) { |
| similarProfileTable = $('#sprofiles').DataTable({ |
| ajax: { |
| method: "GET", |
| url: '/derekovecs/getSimilarProfiles', |
| dataType: 'json', |
| dataSrc: "", |
| timeout: 30000, |
| data: { w: paraResults[0][0].rank } |
| }, |
| "initComplete":function(settings, json){ |
| $('td.paradigmator a').on('mousedown', function(e) { |
| if (e.which === 2) { |
| e.preventDefault(); |
| queryKorAPalternatives(paraResults[0][0].word, this.childNodes["0"].textContent); |
| return false; |
| } |
| }); |
| }, |
| "sScrollY": "780px", |
| "bScrollCollapse": true, |
| "bPaginate": false, |
| "bJQueryUI": true, |
| "dom": '<"top">rt<"bottom"flp><"clear">', |
| "columns": [ |
| { "data": "v", render: function ( data, type, row ) {return data.toFixed(3) }}, |
| { "data": "w", sClass: "paradigmator", render: function ( data, type, row ) {urlprefix.set("word", data); return '<a href="?' + urlprefix + '">' + data + '</a>' } } |
| ], |
| "columnDefs": [ |
| { className: "dt-right", "targets": [0] }, |
| ], |
| "oLanguage": { |
| "sSearch": "Filter: " |
| }, |
| "order": [[ 0, 'desc' ]], |
| }); |
| } |
| // var filterQuot = /(^quot?=[A-Z])|(quot$)/g; |
| var filterQuot = /^quot/; |
| classicCollocatorTable = $('#classicoloctable').DataTable({ |
| ajax: { |
| method: "GET", |
| url: '/derekovecs/getClassicCollocators', |
| dataType: 'json', |
| dataSrc: "", |
| timeout: 30000, |
| data: { w: paraResults[0][0].rank } |
| }, |
| "initComplete":function(settings, json){ |
| $("td.collocator").click(function(){ |
| queryKorAPCII(this.textContent + " /w5 " + CIIsearchWords); |
| }); |
| classicCollocatorTable.columns(".detail").visible(false); |
| $("#ccd").css('width', '450px'); |
| classicCollocatorTable.columns.adjust().draw(); |
| }, |
| "sScrollY": "780px", |
| "bScrollCollapse": true, |
| "bPaginate": false, |
| "bJQueryUI": true, |
| "dom": '<"top">rt<"bottom"flp><"clear">', |
| "columns": [ |
| // { "data": "pos", width: "7%", sClass: "dt-center mono compact", render: function ( data, type, row ) {return bitvec2window(data, row.heat, row.word) }}, |
| { "data": "llr", render: function ( data, type, row ) {return data.toFixed(0) }}, |
| { "data": "lfmd", render: function ( data, type, row ) {return data.toFixed(1) }}, |
| // { "data": "fpmi", type: "scientific", render: function ( data, type, row ) {return data.toExponential(2) } }, |
| { "data": "npmi", render: function ( data, type, row ) {return data.toFixed(2) }}, |
| { "data": "llfmd", render: function ( data, type, row ) {return data.toFixed(1) }}, |
| { "data": "rlfmd", render: function ( data, type, row ) {return data.toFixed(1) }}, |
| { "data": "lnpmi", render: function ( data, type, row ) {return data.toFixed(2) }}, |
| { "data": "rnpmi", render: function ( data, type, row ) {return data.toFixed(2) }}, |
| { "data": "f" }, |
| { "data": "word", sClass: "collocator" } |
| ], |
| "columnDefs": [ |
| { className: "dt-right", "targets": [0,1,2,3,4,5,6,7] }, |
| { className: "dt-right detail", "targets": [3,4,5,6] }, |
| { "searchable": false, |
| "orderable": false, |
| "targets": [] |
| }, |
| { "orderSequence": [ "desc" ], "targets": [ 0, 1, 2,3,4,5,6,7 ] }, |
| { "orderSequence": [ "asc", "desc" ], "targets": [ 8 ] }, |
| ], |
| "oLanguage": { |
| "sSearch": "Filter: " |
| }, |
| "order": [[ 1, 'desc' ]], |
| }); |
| |
| |
| $('#show-details').change(function (e) { |
| var columns = classicCollocatorTable.columns(".detail"); |
| if(this.checked) { |
| columns.visible(true); |
| $("#ccd").css('width', 'auto'); |
| } else { |
| columns.visible(false); |
| $("#ccd").css('width', '450px'); |
| } |
| classicCollocatorTable.columns.adjust().draw(); |
| } ); |
| |
| $("td.collocator").click(function(){ |
| queryKorAPCII(this.textContent + " /w5 " + CIIsearchWords); |
| }); |
| |
| collocatorTable.on( 'order.dt search.dt', function () { |
| collocatorTable.column(0, {order:'applied'}).nodes().each( function (cell, i) { |
| cell.innerHTML = i+1; |
| } ); |
| }).draw(); |
| } |
| |
| if (localStorage && !window.location.hash) { // let's not crash if some user has IE7 |
| var index = parseInt(localStorage['tab']||'0'); |
| $("#tabs").tabs({ active: index }); |
| } |
| $("#tabs").css("visibility", "visible"); // now we can show the tabs |
| }); |
| |
| $(function(){ |
| $("#dropdownoptions").dialog({ |
| title: "Options", |
| autoOpen: false, |
| modal: false, |
| draggable: false, |
| height: "auto", |
| width: "auto", |
| resizable: false, |
| buttons: { |
| "Cancel": function() { |
| $( this ).dialog( "close" ); |
| }, |
| "Apply": function() { |
| window.open($(location).attr('pathname')+'?'+$('form').serialize(), "_self"); |
| } |
| } |
| }); |
| }); |
| |
| $(function(){ |
| $("#showoptions").click(function(){ |
| $("#dropdownoptions").dialog("open"); |
| var target = $(this); |
| $("#dropdownoptions").dialog("widget").position({ |
| my: 'left bottom', |
| at: 'left bottom', |
| of: target |
| }); |
| }); |
| }); |
| |
| $( function() { |
| $( "#no_iterations" ).spinner({ |
| spin: function( event, ui ) { |
| if ( ui.value < 1000 ) { |
| $( this ).spinner( "value", 1000 ); |
| return false; |
| } else if ( ui.value > 10000 ) { |
| $( this ).spinner( "value", 10000 ); |
| return false; |
| } |
| } |
| }); |
| } ); |
| |
| $( function() { |
| $( "#neighbours" ).spinner({ |
| spin: function( event, ui ) { |
| if ( ui.value < 0 ) { |
| $( this ).spinner( "value", 0 ); |
| return false; |
| } else if ( ui.value > 200 ) { |
| $( this ).spinner( "value", 200 ); |
| return false; |
| } |
| } |
| }); |
| } ); |
| |
| $( function() { |
| $( "#cutoff" ).spinner({ |
| spin: function( event, ui ) { |
| if ( ui.value < 100000 ) { |
| $( this ).spinner( "value", 100000 ); |
| return false; |
| } else if ( ui.value > 2000000 ) { |
| $( this ).spinner( "value", 2000000 ); |
| return false; |
| } |
| } |
| }); |
| } ); |
| |
| $( function() { |
| $( "#tabs" ).tabs().addClass('tabs-min'); |
| } ); |
| |
| $( function() { |
| $( ".controlgroup-vertical" ).controlgroup({ |
| "direction": "vertical" |
| }); |
| } ); |
| |
| $(function() { |
| $( document ).tooltip({ |
| content: function() { |
| return $(this).attr('title'); |
| }} |
| ) |
| }); |
| |
| $(function () { |
| $(document).tooltip({ |
| content: function () { |
| return $(this).prop('title'); |
| }, |
| show: null, |
| close: function (event, ui) { |
| ui.tooltip.hover( |
| function () { |
| $(this).stop(true).fadeTo(400, 1); |
| }, |
| function () { |
| $(this).fadeOut("400", function () { |
| $(this).remove(); |
| }) |
| }); |
| } |
| }); |
| }); |
| </script> |
| <script src="//d3js.org/d3.v3.min.js" charset="utf-8"></script> |
| <script src="/derekovecs/js/tsne.js"></script> |
| <script src="/derekovecs/js/som.js"></script> |
| <script src="/derekovecs/js/labeler.js"></script> |
| <script src="/derekovecs/js/derekovcs.js"></script> |
| <script> |
| |
| var opt = {epsilon: <%= $epsilon %>, perplexity: <%= $perplexity %>}, |
| mapWidth = 800, // width map |
| mapHeight = 800, |
| jitterRadius = 7; |
| |
| var T = new tsnejs.tSNE(opt); // create a tSNE instance |
| |
| var Y; |
| |
| var data; |
| var labeler; |
| |
| function applyJitter() { |
| svg.selectAll('.tsnet') |
| .data(labels) |
| .transition() |
| .duration(50) |
| .attr("transform", function(d, i) { |
| T.Y[i][0] = (d.x - mapWidth/2 - tx)/ss/20; |
| T.Y[i][1] = (d.y - mapHeight/2 - ty)/ss/20; |
| return "translate(" + |
| (d.x) + "," + |
| (d.y) + ")"; |
| }); |
| } |
| |
| function updateEmbedding() { |
| var Y = T.getSolution(); |
| svg.selectAll('.tsnet') |
| .data(data.words) |
| .attr("transform", function(d, i) { |
| return "translate(" + |
| ((Y[i][0]*20*ss + tx) + mapWidth/2) + "," + |
| ((Y[i][1]*20*ss + ty) + mapHeight/2) + ")"; }); |
| } |
| |
| var svg; |
| var labels = []; |
| var anchor_array = []; |
| var text; |
| |
| function drawEmbedding() { |
| var urlprefix = new URLSearchParams(window.location.search); |
| urlprefix.delete("word"); |
| urlprefix.append("word",""); |
| |
| $("#embed").empty(); |
| var div = d3.select("#embed"); |
| |
| // get min and max in each column of Y |
| var Y = T.Y; |
| |
| svg = div.append("svg") // svg is global |
| .attr("width", mapWidth) |
| .attr("height", mapHeight); |
| |
| var g = svg.selectAll(".b") |
| .data(data.words) |
| .enter().append("g") |
| .attr("class", "tsnet"); |
| |
| g.append("a") |
| .attr("xlink:href", function(word) { |
| return "?"+urlprefix+word; }) |
| .attr("class", function(d, i) { |
| var res=""; |
| if(data.marked[i]) { |
| res="marked "; |
| } |
| if(data.target.indexOf(" "+d+" ") >= 0) { |
| return res+"target"; |
| } else if(data.ranks[i] < data.mergedEnd) { |
| return res+"merged"; |
| } else { |
| return res; |
| } |
| }) |
| .attr("title", function(d, i) { |
| if(data.mergedEnd > 0) { |
| if(data.ranks[i] >= data.mergedEnd) { |
| return "rank: "+i +" "+"freq. rank: "+(data.ranks[i]).toString().replace(/\B(?=(\d{3})+(?!\d))/g, ","); |
| } else { |
| return "rank: "+i +" "+"freq. rank: "+data.ranks[i].toString().replace(/\B(?=(\d{3})+(?!\d))/g, ",") + " (merged vocab)"; |
| } |
| } else { |
| return "rank: "+i +" "+"freq. rank: "+data.ranks[i].toString().replace(/\B(?=(\d{3})+(?!\d))/g, ","); |
| } |
| }) |
| .append("text") |
| .attr("text-anchor", "top") |
| .attr("font-size", 12) |
| .text(function(d) { return d; }); |
| |
| g.append("svg:title") |
| .text(function(d, i) { |
| return "distance rank: "+ i +" "+"\nfrequency rank: "+data.ranks[i].toString().replace(/\B(?=(\d{3})+(?!\d))/g, ","); |
| }); |
| |
| var zoomListener = d3.behavior.zoom() |
| .scaleExtent([0.1, 10]) |
| .center([0,0]) |
| .on("zoom", zoomHandler); |
| zoomListener(svg); |
| } |
| |
| var tx=0, ty=0; |
| var ss=1; |
| var iter_id=-1; |
| |
| function zoomHandler() { |
| tx = d3.event.translate[0]; |
| ty = d3.event.translate[1]; |
| ss = d3.event.scale; |
| updateEmbedding(); |
| } |
| |
| var stepnum = 0; |
| |
| function stopStep() { |
| clearInterval(iter_id); |
| text = svg.selectAll("text"); |
| |
| // jitter function needs different data and co-ordinate representation |
| labels = d3.range(data.words.length).map(function(i) { |
| var x = (T.Y[i][0]*20*ss + tx) + mapWidth/2; |
| var y = (T.Y[i][1]*20*ss + ty) + mapHeight/2; |
| anchor_array.push({x: x, y: y, r: jitterRadius}); |
| return { |
| x: x, |
| y: y, |
| name: data.words[i] |
| }; |
| }); |
| |
| // get the actual label bounding boxes for the jitter function |
| var index = 0; |
| text.each(function() { |
| labels[index].width = this.getBBox().width; |
| labels[index].height = this.getBBox().height; |
| index += 1; |
| }); |
| |
| |
| // setTimeout(updateEmbedding, 1); |
| // setTimeout( |
| labeler = d3.labeler() |
| .label(labels) |
| .anchor(anchor_array) |
| .width(mapWidth) |
| .height(mapHeight) |
| .update(applyJitter); |
| // .start(1000); |
| |
| iter_id = setInterval(jitterStep, 1); |
| } |
| |
| var jitter_i=0; |
| |
| function jitterStep() { |
| if(jitter_i++ > 100) { |
| clearInterval(iter_id); |
| } else { |
| labeler.start2(10); |
| applyJitter(); |
| } |
| } |
| |
| var last_cost=1000; |
| |
| function step() { |
| var i = T.iter; |
| |
| if(i > <%= $no_iterations %>) { |
| stopStep(); |
| } else { |
| var cost = Math.round(T.step() * 100000) / 100000; // do a few steps |
| $("#cost").html("tsne iteration " + i + ", cost: " + cost.toFixed(5)); |
| if(i % 250 == 0 && cost >= last_cost) { |
| stopStep(); |
| } else { |
| last_cost = cost; |
| updateEmbedding(); |
| } |
| } |
| } |
| |
| function showMap(j) { |
| data=j; |
| T.iter=0; |
| iter_id = -1; |
| last_cost=1000; |
| T.initDataRaw(data.vecs); // init embedding |
| drawEmbedding(); // draw initial embedding |
| |
| if(iter_id >= 0) { |
| clearInterval(iter_id); |
| } |
| //T.debugGrad(); |
| iter_id = setInterval(step, 1); |
| if(true) { // (<%= $show_som %>) { |
| makeSOM(j, <%= $no_iterations %>); |
| } |
| } |
| var queryword; |
| |
| function showCollocatorSOM() { |
| if (collocatorTable) { |
| var ctableData = collocatorTable.rows().data(); |
| var nwords = [], |
| nranks = []; |
| for (var i=0; i < ctableData.length && i < 100; i++) { |
| nranks.push(ctableData[i].rank); |
| nwords.push(ctableData[i].word); |
| } |
| $.post('/derekovecs/getVecsByRanks', |
| JSON.stringify(nranks), |
| function(data, status){ |
| showMap({target: " "+urlParams.get('word')+" ", mergedEnd: 0, words: nwords, vecs: data, ranks: nranks, marked: Array(100).fill(false)} ); |
| }, 'json'); |
| } |
| } |
| |
| function onload() { |
| queryword = document.getElementById('word'); |
| } |
| |
| function queryKorAP() { |
| window.open('http://korap.ids-mannheim.de/kalamar/?q='+queryword.value, 'KorAP'); |
| } |
| |
| function queryKorAPCII(query) { |
| window.open('http://korap.ids-mannheim.de/kalamar/?ql=cosmas2&q='+query, 'KorAP'); |
| } |
| </script> |
| </head> |
| <body onload="onload()"> |
| <div style="display:none;" id="lfmd_ttt">PMI cubed [1], also called log-frequency biased mutual dependency [2]: $$\text{PMI}^3=\text{LFMD}=log_2\frac{p^3(w_1,w_2)}{p(w_1) p(w_2)}$$<p class="citation">[1] Daille, B. (1994): <a href="http://www.bdaille.com/index.php?option=com_docman&task=doc_download&gid=8&Itemid=">Approche mixte pour l’extraction automatique de terminologie: statistiques lexicales et filtres linguistiques</a>. PhD thesis, Université Paris 7.</p><p class="citation">[2] Thanopoulos, A., Fakotakis, N., Kokkinakis, G. (2002): <a href="http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.11.8101&rep=rep1&type=pdf">Comparative evaluation of collocation extraction metrics</a>. In: Proc. of LREC 2002: 620–625.</p></div> |
| <div style="display:none;" id="npmi_ttt">Normalized pointwise mutual information: $$\frac{log_2\frac{p(w_1,w_2)}{p(w_1)p(w_2)}}{-log_2(p(w_1,w_2))}$$<p class="citation">Bouma, Gerlof (2009): <a href="https://svn.spraakdata.gu.se/repos/gerlof/pub/www/Docs/npmi-pfd.pdf">Normalized (pointwise) mutual information in collocation extraction</a>. In Proceedings of GSCL.</p></div> |
| <div id="header"> |
| <div id="pagetitle"> |
| <h1>DeReKo-Vectors</h1> |
| </div> |
| <div id="options" class="widget"> |
| <form id="queryform"> |
| <input id="word" type="text" name="word" placeholder="Word(s) to be searched" value="<%= $word %>" |
| title="When looking for multiple words use spaces as separators to search around the average vector and | as separator to get the neighbours for each word."/> |
| <input id="SEARCH" type="button" value="SEARCH"> |
| <input type="button" id="showoptions" name="showoptions" value="Options" /> |
| </form> |
| <div id="dropdownoptions" style="display: none"> |
| <form id="optionsform"> |
| <div class="controlgroup-vertical"> |
| <label for="cutoff">cut-off</label> |
| <input id="cutoff" type="text" name="cutoff" size="10" value="<%= $cutoff %>" title="Only consider the most frequent x word forms."> |
| <label for="dedupe">dedupe</label> |
| <input id="dedupe" type="checkbox" name="dedupe" value="1" <%= ($dedupe ? "checked" : "") %> title="radically filter out any near-duplicates"> |
| % if($mergedEnd > 0) { |
| <label for="sbf">backw.</label> |
| <input id="sbf" type="checkbox" name="sbf" value="1" <%= ($searchBaseVocabFirst ? "checked" : "") %> title="If checkecked base vocabulary will be searched first. Otherwise merged vocabulray will be searched first."> |
| % } |
| <label for="neighbours">max. neighbours:</label> |
| <input id="neighbours" size="4" name="n" value="<%= $no_nbs %>"> |
| <label for="no_iterations">max. iterations</label> |
| <input id="no_iterations" name="N" size="4" value="<%= $no_iterations %>"> |
| <!-- <label for="dosom">SOM</label> |
| <input id="dosom" type="checkbox" name="som" value="1" <%= ($show_som ? "checked" : "") %>> --> |
| % if($collocators) { |
| <label for="sortby">window/sort</label> |
| <select id="sortby" name="sort"> |
| <option value="0" <%= ($sort!=1 && $sort!=2? "selected":"") %>>auto focus</option> |
| <!-- <option value="1" <%= ($sort==1? "selected":"") %>>any single position</option> |
| <option value="2" <%= ($sort==2? "selected":"") %>>whole window</option> --> |
| </select> |
| % } |
| <input type="button" value="→ KorAP" onclick="queryKorAP();" title="query word with KorAP"/> |
| <input id="show-details" type="checkbox" name="show-details" value="1" > |
| <label for="show-details"> |
| Show details |
| </label> |
| </div> |
| </form> |
| </div> |
| </div> |
| </div> |
| <div id="topwrapper"> |
| <div style="visibility: hidden;" id="tabs"> |
| <ul> |
| <li><a href="#tabs-1">Semantics (TSNE-map)</a></li> |
| <li><a href="#tabs-2">Semantics (SOM)</a></li> |
| <li><a href="#tabs-3">Syntagmatic (collocators)</a></li> |
| </ul> |
| <div id="tabs-1" style="display: flex; padding: 5px; flex-flow: row wrap;"> |
| % if($lists && (@$lists) > 0 && (@$lists)[0]) { |
| <div id="wrapper"> |
| <div id="first" style="width: 230px; margin-bottom: 15px;"> |
| <table class="display compact nowrap" id="firsttable"> |
| <thead> |
| <tr> |
| <th align="right">#</th><th align="right">cos</th><th align="left">similars by w2v</th> |
| </tr> |
| </thead> |
| <tbody> |
| <tr> |
| <td align="right"> |
| </td> |
| <td align="right"> |
| </td> |
| <td></td> |
| </tr> |
| </tbody> |
| </table> |
| </div> |
| % if($haveSProfiles) { |
| <div id="sprofilesdiv" style="width: 200px; padding-right: 10px;"> |
| <table class="display compact nowrap" id="sprofiles"> |
| <thead> |
| <tr> |
| <th align="right">cos</th><th align="left">similars by coll. profile</th> |
| </tr> |
| </thead> |
| <tbody> |
| <tr> |
| <td align="right"> |
| </td> |
| <td></td> |
| </tr> |
| </tbody> |
| </table> |
| </div> |
| %} |
| <div id="second"> |
| <div id="embed"> |
| </div> |
| <div id="cost"> |
| </div> |
| </div> |
| </div> |
| % } elsif($word !~ /^\s*$/) { |
| <div id="wrapper"> |
| <div id="not-found-dialog" title="Not found"> |
| <p>ERROR: "<%= $word %>" not found in vocabluary.</p> |
| <p>If you are sure you have spelled the word as intended, you can try to increase the cutoff parameter in the options menu.</p> |
| </div> |
| <script> |
| $( function() { |
| $( "#not-found-dialog" ).dialog({ |
| autoOpen: true, |
| modal: true, |
| draggable: false, |
| height: "auto", |
| width: "auto", |
| resizable: false, |
| buttons: { |
| "OK": function() { |
| $( this ).dialog( "close" ); |
| }, |
| "Apply": function() { |
| window.open($(location).attr('pathname')+'?'+$('form').serialize(), "_self"); |
| } |
| } |
| }); |
| }); |
| </script> |
| </div> |
| % } |
| </div> |
| <div id="tabs-2" style="display: flex; padding: 5px; flex-flow: row wrap;"> |
| <div id="som2" style="width: 800;"> |
| <div id="sominfo1"><span id="somcolor1"> </span> <span id="somword1"> </span> <span id="somcolor2"> </span> <span id="somword2"> </span> <span id="somcolor3"> </span></div> |
| <div id="sominfo" style="text-align: right">SOM iteration <span id="iterations">0</span></div> |
| </div> |
| </div> |
| <div id="tabs-3" style="display: flex; padding:5px; flex-flow: row wrap;"> |
| <div style="margin-right: 20px; margin-bottom: 10px;" id="secondt"> |
| <table class="display compact nowrap" id="secondtable"> |
| <thead> |
| <tr> |
| % if($collocators) { |
| <th>#</th> |
| <th align="center" title="Activation of the respective collocator in the columns around the target normalized by its maximum (red). Columns selected by the auto-focus funtion (which window of all possible column-combinations maximizes ⊥(a/c)?) are marked with +. Click on the column postions to lauch a KorAP query with target word and collocator in the respective position.">w'</th> |
| <th align="right" title="Maximum activation of the collocator anywhere in the output layer.">max(a)</th> |
| <th title="Average raw activation of the collocator in the columns selected by auto-focus." align="right">⟨a⟩</th> |
| <th title="Sum of activations over the selected colunns normalized by the total activation sum of the selected columns." align="right">Σa/Σw'</th> |
| <th title="Co-norm of the column-normalized activations over the colunns selected by the auto-focus." align="right">⊥(a/c)</th> |
| <th title="Sum of the activations over the whole window normalized by the total window sum (no auto-focus)." align="right">Σa/Σw</th> |
| <th align="left">collocator (by w2v)</th> |
| % } |
| </tr> |
| </thead> |
| <tbody> |
| <tr> |
| <td align="right"> |
| </td> |
| <td align="right"> |
| </td> |
| <td align="right"> |
| </td> |
| <td align="right"> |
| </td> |
| <td align="right"> |
| </td> |
| <td align="right"> |
| </td> |
| </tr> |
| </tbody> |
| </table> |
| </div> |
| <div id="ccd" style=""> |
| <table class="display compact nowrap" id="classicoloctable"> |
| <thead> |
| % if($collocators) { |
| <tr> |
| <th>LLR</th> |
| <th id="lfmd_tt">PMI³</th> |
| <th id="npmi_tt">nPMI</th> |
| <th title="PMI³ restricted to left neighbour">l-PMI³</th> |
| <th title="PMI³ restricted to right neighbour">r-PMI³</th> |
| <th title="nPMI restricted to left neighbour">l-nPMI</th> |
| <th title="nPMI restricted to right neighbour">r-nPMI</th> |
| <th title="raw max frequency of collocation within window">raw</th> |
| <th>collocator (by collocation analysis)</th> |
| </tr> |
| % } |
| </thead> |
| <tbody> |
| <tr> |
| <td align="right"> |
| </td> |
| <td align="right"> |
| </td> |
| <td align="right"> |
| </td> |
| <td align="right"> |
| </td> |
| <td align="right"> |
| </td> |
| <td align="right"> |
| </td> |
| <td align="right"></td> |
| <td align="right"></td> |
| <td align="left"></td> |
| </tr> |
| </tbody> |
| </table> |
| </div> |
| <!-- |
| <div style="clear:both" ></div> |
| <div style="float: right; overflow: hidden" id="extra"><button onClick="showCollocatorSOM()"> </button></div> |
| --> |
| </div> |
| </div> |
| </div> <!-- topwrapper --> |
| <div style="clear: both;"></div> |
| </div> |
| % if($training_args && (@$lists)[0]) { |
| <div class="info"> |
| % if($training_args =~ /-type\s*3/) { |
| Calculations are based on a word embedding model trained with an extension of <a href="https://github.com/wlin12/wang2vec/">wang2vec</a> using the following parameters: <%= $training_args %> |
| % } else { |
| Calculations are based on a word embedding model trained with <a href="https://code.google.com/p/word2vec/">word2vec</a> using the following parameters: <%= $training_args %> |
| % } |
| </div> |
| % } |
| </body> |
| </html> |