~~This paper presents empirical studies and closely corresponding theoretical models ofthe performance of a chart parser exhaus tively parsing the Penn Treebank with the Treebank?s own CFG grammar.~~
~~We show how performance is dramatically affected by rule representation and tree transformations,but little by top-down vs. bottom-up strategies.~~
~~We discuss grammatical saturation, in cluding analysis of the strongly connected components of the phrasal nonterminals in the Treebank, and model how, as sentence length increases, the effective grammar rule size increases as regions of the grammar are unlocked, yielding super-cubic observed time behavior in some configurations.~~
~~This paper originated from examining the empiricalperformance of an exhaustive active chart parser us ing an untransformed treebank grammar over the PennTreebank.~~
~~Our initial experiments yielded the sur prising result that for many configurations empirical parsing speed was super-cubic in the sentence length.~~
~~This led us to look more closely at the structure of the treebank grammar.~~
~~The resulting analysis builds on the presentation of Charniak (1996), but extendsit by elucidating the structure of non-terminal inter relationships in the Penn Treebank grammar.~~
~~On the basis of these studies, we build simple theoreticalmodels which closely predict observed parser performance, and, in particular, explain the originally ob served super-cubic behavior.~~
~~We used treebank grammars induced directly from the local trees of the entire WSJ section of the Penn Treebank (Marcus et al, 1993) (release 3).~~
~~For eachlength and parameter setting, 25 sentences evenly dis tributed through the treebank were parsed.~~
~~Since we were parsing sentences from among those from whichour grammar was derived, coverage was never an is sue.~~
~~Every sentence parsed had at least one parse ? the parse with which it was originally observed.1The sentences were parsed using an implementation of the probabilistic chart-parsing algorithm pre sented in (Klein and Manning, 2001).~~
~~In that paper, we present a theoretical analysis showing an worst-case time bound for exhaustively parsing arbi trary context-free grammars.~~
~~In what follows, we donot make use of the probabilistic aspects of the gram mar or parser.~~
~~The parameters we varied were: Tree Transforms: NOTRANSFORM, NOEMPTIES, NOUNARIESHIGH, and NOUNARIESLOW.~~
~~Grammar Rule Encodings: LIST, TRIE, or MIN Rule Introduction: TOPDOWN or BOTTOMUP The default settings are shown above in bold face.~~
~~We do not discuss all possible combinations of these settings.~~
~~Rather, we take the bottom-up parser using an untransformed grammar with trie rule encodings to be the basic form of the parser.~~
~~Except where noted, we will discuss how each factor affects this baseline, as most of the effects are orthogonal.~~
~~When we name a setting, any omitted parameters are assumed to be the defaults.~~
~~2.1 Tree Transforms.~~
~~In all cases, the grammar was directly induced from (transformed) Penn treebank trees.~~
~~The transformsused are shown in figure 1.~~
~~For all settings, func tional tags and crossreferencing annotations were stripped.~~
~~For NOTRANSFORM, no other modification was made.~~
~~In particular, empty nodes (represented as -NONE- in the treebank) were turned into rules thatgenerated the empty string ( ), and there was no collapsing of categories (such as PRT and ADVP) as is of ten done in parsing work (Collins, 1997, etc.).~~
~~For 1Effectively ?testing on the training set?~~
~~would be invalid if we wished to present performance results such as precision and recall, but it is not a problem for the present experiments, which focus solely on the parser load and grammar structure.~~
~~TOP S-HLN NP-SBJ-NONE VP VB Atone TOP S NP-NONE VP VB Atone TOP S VP VB Atone TOP S Atone TOP VB Atone (a) (b) (c) (d) (e)Figure 1: Tree Transforms: (a) The raw tree, (b) NOTRANSFORM, (c) NOEMPTIES, (d) NOUNARIES HIGH (e) NOUNARIESLOW NNP NNS NNP NNP NNJJ NNS JJ CD NNCD NNDT NNDT NN NNS DT JJ DT NN CC NP NP PP NP SBAR NP NNSNN NN PRP QP NNS NNS NNS NNS NNP NNP NN JJ CD NN NN DT NN JJ NN NP CC NP NN SBAR PP PRP QP NNS NNS NNS NNP NNP JJ NN CD NN NNS DT JJ NN NP NP CC PP SBAR NN PRP QP NN LIST TRIE MIN Figure 2: Grammar Encodings: FSAs for a subset of the rules for the category NP.~~
~~Non-black states areactive, non-white states are accepting, and bold transi tions are phrasal.NOEMPTIES, empties were removed by pruning nonterminals which covered no overt words.~~
~~For NOUNA RIESHIGH, and NOUNARIESLOW, unary nodes wereremoved as well, by keeping only the tops and the bot toms of unary chains, respectively.2 2.2 Grammar Rule Encodings.~~
~~The parser operates on Finite State Automata (FSA) grammar representations.~~
~~We compiled grammar rules into FSAs in three ways: LISTs, TRIEs, andMINimized FSAs.~~
~~An example of each representa tion is given in figure 2.~~
~~For LIST encodings, eachlocal tree type was encoded in its own, linearly structured FSA, corresponding to Earley (1970)-style dotted rules.~~
~~For TRIE, there was one FSA per cate gory, encoding together all rule types producing thatcategory.~~
~~For MIN, state-minimized FSAs were con structed from the trie FSAs.~~
~~Note that while the rule encoding may dramatically affect the efficiency of a parser, it does not change the actual set of parses for a given sentence in any way.32In no case were the nonterminal-to-word or TOP-to nonterminal unaries altered.3FSAs are not the only method of representing and com pacting grammars.~~
~~For example, the prefix compacted tries we use are the same as the common practice of ignoring items before the dot in a dotted rule (Moore, 2000).~~
~~Another 0 60 120 180 240 300 360 0 10 20 30 40 50 SentenceLength A vg .T im e(s ec on ds ) List-NoTransform exp3.54r0.999 Trie-NoTransform exp3.16r0.995 Trie-NoEmpties exp3.47r0.998 Trie-NoUnariesHigh exp3.67r0.999 Trie-NoUnariesLow exp3.65r0.999 Min-NoTransform exp2.87r0.998 Min-NoUnariesLow exp3.32r1.000 Figure 3: The average time to parse sentences using various parameters.~~
~~In this section, we outline the observed performance of the parser for various settings.~~
~~We frequently speak in terms of the following: span: a range of words in the chart, e.g., [1,3]4 edge: a category over a span, e.g., NP:[1,3] traversal: a way of making an edge from an active and a passive edge, e.g., NP:[1,3] (NP DT.NN:[1,2] + NN:[2,3]) 3.1 Time.~~
~~The parser has an theoretical time bound, where is the number of words in the sentence to be parsed, is the number of nonterminal categories in the grammar and is the number of (active) states in the FSA encoding of the grammar.~~
~~The time boundis derived from counting the number of traversals pro cessed by the parser, each taking time.In figure 3, we see the average time5 taken per sentence length for several settings, with the empirical exponent (and correlation -value) from the best-fit sim ple power law model to the right.~~
~~Notice that most settings show time growth greater than . Although, is simply an asymptotic bound,there are good explanations for the observed behav ior.~~
~~There are two primary causes for the super-cubic time values.~~
~~The first is theoretically uninteresting.~~
~~The parser is implemented in Java, which uses garbage collection for memory management.~~
~~Even when thereis plenty of memory for a parse?s primary data struc tures, ?garbage collection thrashing?~~
~~can occur when logical possibility would be trie encodings which compact the grammar states by common suffix rather than common prefix, as in (Leermakers, 1992).~~
~~The savings are less than for prefix compaction.~~
~~4Note that the number of words (or size) of a span is equal to the difference between the endpoints.~~
~~5The hardware was a 700 MHz Intel Pentium III, and we used up to 2GB of RAM for very long sentences or very poor parameters.~~
~~With good parameter settings, the system can parse 100+ word treebank sentences.~~
~~0.0M 5.0M 10.0M 15.0M 20.0M 0 10 20 30 40 50 SentenceLength A vg .T ra ve rs al s NoTransform exp2.86r1.000 NoEmpties exp3.28r1.000 NoUnariesHigh exp3.74r0.999 NoUnariesLow exp3.83r0.999 0.0M 5.0M 10.0M 15.0M 20.0M 0 10 20 30 40 50 SentenceLength A vg .T ra ve rs al s List exp2.60r0.999 Trie exp2.86r1.000 Min exp2.78r1.000 0.994 0.995 0.996 0.997 0.998 0.999 1.000 1.001 1.002 0 10 20 30 40 50 SentenceLength R at io (T D/ BU ) Edges Traversals (a) (b) (c) Figure 4: (a) The number of traversals for different grammar transforms.~~
~~(b) The number of traversals for different grammar encodings.~~
~~(c) The ratio of the number of edges and traversals produced with a top-down strategy over the number produced with a bottom-up strategy (shown for TRIE-NOTRANSFORM, others are similar).~~
~~parsing longer sentences as temporary objects causeincreasingly frequent reclamation.~~
~~To see past this ef fect, which inflates the empirical exponents, we turn to the actual traversal counts, which better illuminate the issues at hand.~~
~~Figures 4 (a) and (b) show the traversal curves corresponding to the times in figure 3.~~
~~The interesting cause of the varying exponents comes from the ?constant?~~
~~terms in the theoretical bound.~~
~~The second half of this paper shows how modeling growth in these terms can accurately predict parsing performance (see figures 9 to 13).~~
~~3.2 Memory.~~
~~The memory bound for the parser is ff . Sincethe parser is running in a garbage-collected environ ment, it is hard to distinguish required memory from utilized memory.~~
~~However, unlike time and traversals which in practice can diverge, memory requirements match the number of edges in the chart almost exactly, since the large data structures are all proportional in size to the number of edges fiffiflff .6 Almost all edges stored are active edges ( "!$#&% for sentences longer than 30 words), of which there can be 'ff : one for every grammar state and span.~~
~~Passive edges, of which there can be 'ff( , one for ev ery category and span, are a shrinking minority.~~
~~Thisis because, while is bounded above by 27 in the tree bank7 (for spans 2), numbers in the thousands (see figure 12).~~
~~Thus, required memory will be implicitly modeled when we model active edges in section 4.3.~~
~~3.3 Tree Transforms.~~
~~Figure 4 (a) shows the effect of the tree transforms on traversal counts.~~
~~The NOUNARIES settings are much more efficient than the others, however this efficiency comes at a price in terms of the utility of the final parse.~~
~~For example, regardless of which NOUNARIES6A standard chart parser might conceivably require storing more than )+*,.- traversals on its agenda, but ours prov ably never does.~~
~~7This count is the number of phrasal categories with the introduction of a TOP label for the unlabeled top treebank nodes.~~
~~transform is chosen, there will be NP nodes missing from the parses, making the parses less useful for any task requiring NP identification.~~
~~For the remainder ofthe paper, we will focus on the settings NOTRANS FORM and NOEMPTIES.~~
~~3.4 Grammar Encodings.~~
~~Figure 4 (b) shows the effect of each tree transform on traversal counts.~~
~~The more compacted the grammar representation, the more time-efficient the parser is. 3.5 Top-Down vs. Bottom-Up.~~
~~Figure 4 (c) shows the effect on total edges and traversals of using top-down and bottom-up strategies.There are some extremely minimal savings in traversals due to top-down filtering effects, but there is a cor responding penalty in edges as rules whose left-cornercannot be built are introduced.~~
~~Given the highly unre strictive nature of the treebank grammar, it is not very surprising that top-down filtering provides such little benefit.~~
~~However, this is a useful observation about real world parsing performance.~~
~~The advantages of top-down chart parsing in providing grammar-driven prediction are often advanced (e.g., Allen 1995:66), but in practice we find almost no value in this for broad coverage CFGs.~~
~~While some part of this is perhaps due to errors in the treebank, a large part just reflects the true nature of broad coverage grammars: e.g., onceyou allow adverbial phrases almost anywhere and al low PPs, (participial) VPs, and (temporal) NPs to beadverbial phrases, along with phrases headed by ad verbs, then there is very little useful top-down control left.~~
~~With such a permissive grammar, the only real constraints are in the POS tags which anchor the local trees (see section 4.3).~~
~~Therefore, for the remainder of the paper, we consider only bottom-up settings.~~
~~In the remainder of the paper we provide simple mod els that nevertheless accurately capture the varying magnitudes and exponents seen for different grammar encodings and tree transformations.~~
~~Since the term of comes directly from the number of start, split, and end points for traversals, it is certainly not responsible for the varying growth rates.~~
~~An initially plausible possibility is that the quantity bounded by the term is non-constant in in practice, becauselonger spans are more ambiguous in terms of the num ber of categories they can form.~~
~~This turns out tobe generally false, as discussed in section 4.2.~~
~~Alter nately, the effective term could be growing with , which turns out to be true, as discussed in section 4.3.The number of (possibly zero-size) spans for a sen tence of length is fixed: 0/1 23/546 8764 . Thus, to be able to evaluate and model the total edge counts, we look to the number of edges over a given span.~~
~~Definition 1 The passive (or active) saturation of a given span is the number of passive (or active) edges over that span.~~
~~In the total time and traversal bound , theeffective value of is determined by the active satu ration, while the effective value of is determined by the passive saturation.~~
~~An interesting fact is that the saturation of a span is, for the treebank grammar andsentences, essentially independent of what size sen tence the span is from and where in the sentence the span begins.~~
~~Thus, for a given span size, we report the average over all spans of that size occurring anywhere in any sentence parsed.~~
~~4.1 Treebank Grammar Structure.~~
~~The reason that effective growth is not found in the component is that passive saturation stays almostconstant as span size increases.~~
~~However, the more interesting result is not that saturation is relatively con stant (for spans beyond a small, grammar-dependent size), but that the saturation values are extremely largecompared to (see section 4.2).~~
~~For the NOTRANS FORM and NOEMPTIES grammars, most categories are reachable from most other categories using rules which can be applied over a single span.~~
~~Once you get one of these categories over a span, you will get the rest as well.~~
~~We now formalize this.~~
~~Definition 2 A category 9 is empty-reachable in agrammar : if 9 can be built using only empty ter minals.~~
~~The empty-reachable set for the NOTRANSFORM grammar is shown in figure 5.8 These 23 categories plus the tag -NONE- create a passive saturation of 24 for zero-spans for NOTRANSFORM (see figure 9).~~
~~Definition 3 A category ; is same-span-reachable from a category 9 in a grammar : if ; can be built from 9 using a parse tree in which, aside from at most8The set of phrasal categories used in the Penn Tree bank is documented in Manning and Schu?tze (1999, 413); Marcus et al (1993, 281) has an early version.~~
~~ADJP ADVP FRAG INTJ NAC NP NX PP PRN QP RRC S SBAR SBARQ SINV SQ TOP UCP VP WHADVP WHNP WHPP XFigure 5: The empty-reachable set for the NOTRANS FORM grammar.~~
~~ADJP ADVP FRAG INTJ NAC NP NX PP PRN QP RRC S SBAR SBARQ SINV SQ UCP VP WHNP X TOP CONJP LST PRT WHADJP WHADVP WHPP Figure 6: The same-span reachability graph for the NOTRANSFORM grammar.~~
~~ADJP ADVP FRAG INTJ NP PP PRN QP S SBAR UCP VP WHNP TOP CONJP LST NAC NX SQ X RRC PRT WHADJP SBARQ WHADVP SINV WHPP Figure 7: The same-span-reachability graph for the NOEMPTIES grammar.one instance of 9 , every node not dominating that in stance is an instance of an empty-reachable category.~~
~~The same-span-reachability relation induces a graphover the 27 non-terminal categories.~~
~~The strongly connected component (SCC) reduction of that graph is shown in figures 6 and 7.9 Unsurprisingly, the largest SCC, which contains most ?common?~~
~~categories (S,NP, VP, PP, etc.) is slightly larger for the NOTRANSFORM grammar, since the empty-reachable set is non empty.~~
~~However, note that even for NOTRANSFORM, the largest SCC is smaller than the empty-reachable set, since empties provide direct entry into some of the lower SCCs, in particular because of WH-gaps.Interestingly, this same high-reachability effect oc curs even for the NOUNARIES grammars, as shown in the next section.~~
~~4.2 Passive Edges.~~
~~The total growth and saturation of passive edges is relatively easy to describe.~~
~~Figure 8 shows the total num 9Implied arcs have been removed for clarity.~~
~~The relation is in fact the transitive closure of this graph.~~
~~0.0K 5.0K 10.0K 15.0K 20.0K 25.0K 30.0K 0 10 20 30 40 50 SentenceLength A vg .P as si ve To ta ls NoTransform exp1.84r1.000 NoEmpties exp1.97r1.000 NoUnariesHigh exp2.13r1.000 NoUnariesLow exp2.21r0.999 0.0K 5.0K 10.0K 15.0K 20.0K 25.0K 30.0K 0 10 20 30 40 50 SentenceLength A vg .P as si ve To ta ls NoTransform exp1.84r1.000 NoEmpties exp1.95r1.000 NoUnariesHigh exp2.08r1.000 NoUnariesLow exp2.20r1.000 Figure 8: The average number of passive edges processed in practice (left), and predicted by our models (right).~~
~~0 5 10 15 20 25 30 0 1 2 3 4 5 6 7 8 9 10 SpanSize A vg .P as si ve Sa tu ra tio n NoTransform NoEmpties NoUnariesHigh NoUnariesLow 0 5 10 15 20 25 30 0 1 2 3 4 5 6 7 8 9 10 SpanSize A vg .P as si ve Sa tu ra tio n NoTransform NoEmpties NoUnariesHigh NoUnariesLow Figure 9: The average passive saturation (number of passive edges) for a span of a given size as processed in practice (left), and as predicted by our models (right).~~
~~ber of passive edges by sentence length, and figure 9 shows the saturation as a function of span size.10 The grammar representation does not affect which passive edges will occur for a given span.~~
~~The large SCCs cause the relative independence ofpassive saturation from span size for the NOTRANS FORM and NOEMPTIES settings.~~
~~Once any category inthe SCC is found, all will be found, as well as all cate gories reachable from that SCC.~~
~~For these settings, thepassive saturation can be summarized by three satura tion numbers: zero-spans (empties) <>=?$@BA , one-spans (words) <>=?$@8C , and all larger spans (categories) <>=?$@ ff . Taking averages directly from the data, we have our first model, shown on the right in figure 9.~~
~~For the NOUNARIES settings, there will be no same-span reachability and hence no SCCs.~~
~~To reach a new category always requires the use of at least one overt word.~~
~~However, for spans of size 6 or so, enough words exist that the same high saturation effect will still be observed.~~
~~This can be modeled quite simply by assuming each terminal unlocks a fixed fraction of the nonterminals, as seen in the right graph of figure 9, but we omit the details here.Using these passive saturation models, we can directly estimate the total passive edge counts by sum mation: <D@FE @2G fl5HIJLK A M/NPORQF S<>=?$@ J 10The maximum possible passive saturation for any span greater than one is equal to the number of phrasal categories in the treebank grammar: 27.~~
~~However, empty and size-one spans can additionally be covered by POS tag edges.The predictions are shown in figure 8.~~
~~For the NO TRANSFORM or NOEMPTIES settings, this reduces to: <T@FE @2G flVU IXW CY I ff <>=Z?&@ ff /NG <[=?&@8C\/]/^ S<[=?&@BAWe correctly predict that the passive edge total ex ponents will be slightly less than 2.0 when unaries are present, and greater than 2.0 when they are not.~~
~~With unaries, the linear terms in the reduced equation are significant over these sentence lengths and drag downthe exponent.~~
~~The linear terms are larger for NO TRANSFORM and therefore drag the exponent downmore.11 Without unaries, the more gradual satura tion growth increases the total exponent, more so for NOUNARIESLOW than NOUNARIESHIGH.~~
~~However, note that for spans around 8 and onward, the saturation curves are essentially constant for all settings.~~
~~4.3 Active Edges.~~
~~Active edges are the vast majority of edges and essen tially determine (non-transient) memory requirements.~~
~~While passive counts depend only on the grammartransform, active counts depend primarily on the en coding for general magnitude but also on the transform for the details (and exponent effects).~~
~~Figure 10 showsthe total active edges by sentence size for three set tings chosen to illustrate the main effects.~~
~~Total active growth is sub-quadratic for LIST, but has an exponent of up to about 2.4 for the TRIE settings.~~
~~11Note that, over these values of _ , even a basic quadratic function like the simple sum H_`^*S_badce-_[fZg has a best fit simple power curve exponent of only hicZj k l for the samereason.~~
~~Moreover, note that _Tmefg has a higher best-fit expo nent, yet will never actually outgrow it.~~
~~0.0M 0.5M 1.0M 1.5M 2.0M 0 10 20 30 40 50 SentenceLength A vg .A ct iv eT ot al s List-NoTransform exp1.88r0.999 Trie-NoTransform exp2.18r0.999 Trie-NoEmpties exp2.43r0.999 0.0M 0.5M 1.0M 1.5M 2.0M 0 10 20 30 40 50 SentenceLength A vg .A ct iv eT ot al s List-NoTransform exp1.81r0.999 Trie-NoTransform exp2.10r1.000 Trie-NoEmpties exp2.36r1.000 Figure 10: The average number of active edges for sentences of a given length as observed in practice (left), and as predicted by our models (right).~~
~~0.0K 2.0K 4.0K 6.0K 8.0K 10.0K 12.0K 14.0K 0 5 10 15 20 SpanLength A vg .A ct iv eS at ur at io n List-NoTransform exp0.092r0.957 Trie-NoTransform exp0.323r0.999 Trie-NoEmpties exp0.389r0.997 0.0K 2.0K 4.0K 6.0K 8.0K 10.0K 12.0K 14.0K 0 5 10 15 20 SpanLength A vg .A ct iv eS at ur at io n List-NoTransform exp0.111r0.999 Trie-NoTransform exp0.297r0.998 Trie-NoEmpties exp0.298r0.991 Figure 11: The average active saturation (number of active edges) for a span of a given size as processed in practice (left), and as predicted by our models (right).~~
~~NOTRANS NOEMPTIES NOUHIGH NOULOW LIST 80120 78233 81287 100818 TRIE 17298 17011 17778 22026 MIN 2631 2610 2817 3250 Figure 12: Grammar sizes: active state counts.To model the active totals, we again begin by mod eling the active saturation curves, shown in figure 11.~~
~~The active saturation for any span is bounded above by , the number of active grammar states (states in the grammar FSAs which correspond to active edges).~~
~~For list grammars, this number is the sum of the lengths of all rules in the grammar.~~
~~For trie grammars, it is the number of unique rule prefixes (including the LHS) in the grammar.~~
~~For minimized grammars, it is the number of states with outgoing transitions (non-black states in figure 2).~~
~~The value of is shown for each setting in figure 12.~~
~~Note that the maximum number ofactive states is dramatically larger for lists since common rule prefixes are duplicated many times.~~
~~For min imized FSAs, the state reduction is even greater.~~
~~Since states which are earlier in a rule are much more likely to match a span, the fact that tries (and min FSAs) compress early states is particularly advantageous.~~
~~Unlike passive saturation, which was relatively close to its bound , active saturation is much farther below . Furthermore, while passive saturation was relatively constant in span size, at least after a point, active saturation quite clearly grows with span size, even for spans well beyond those shown in figure 11.~~
~~We now model these active saturation curves.~~
~~What does it take for a given active state to match agiven span?~~
~~For TRIE and LIST, an active state cor responds to a prefix of a rule and is a mix of POS tags and phrasal categories, each of which must be matched, in order, over that span for that state to be reached.~~
~~Given the large SCCs seen in section 4.1, phrasal categories, to a first approximation, might as well be wildcards, able to match any span, especiallyif empties are present.~~
~~However, the tags are, in com parison, very restricted.~~
~~Tags must actually match a word in the span.~~
~~More precisely, consider an active state ? in thegrammar and a span = . In the TRIE and LIST encod ings, there is some, possibly empty, list n of labels that must be matched over = before an active edge with this state can be constructed over that span.12 Assume that the phrasal categories in n can match any span (or any non-zero span in NOEMPTIES).13 Therefore, phrasal categories in n do not constrain whether ? can match = . The real issue is whether the tags in n will match words in = . Assume that a random tag matches a random word with a fixed probability < , independently of where the tag is in the rule and where the word is inthe sentence.14 Assume further that, although tags oc cur more often than categories in rules (63.9% of rule items are tags in the NOTRANSFORM case15), given a 12The essence of the MIN model, which is omitted here, is that states are represented by the ?easiest?~~
~~label sequence which leads to that state.~~
~~13The model for the NOUNARIES cases is slightly more complex, but similar.~~
~~14This is of course false; in particular, tags at the end of rules disproportionately tend to be punctuation tags.~~
~~15Although the present model does not directly apply to the NOUNARIES cases, NOUNARIESLOW is significantly fixed number of tags and categories, all permutations are equally likely to appear as rules.16Under these assumptions, the probability that an ac tive state ? is in the treebank grammar will depend only on the number @ of tags and o of categories in n . Call this pair p\?q rfls@et8oZ the signature of ? .~~
~~For a given signature p , let o2Eu[v@2p be the number of ac tive states in the grammar which have that signature.~~
~~Now, take a state ? of signature @etwox and a span = . If we align the tags in ? with words in = and alignthe categories in ? with spans of words in = , then pro vided the categories align with a non-empty span (for NOEMPTIES) or any span at all (for NOTRANSFORM), then the question of whether this alignment of ? with = matches is determined entirely by the @ tags.~~
~~However, with our assumptions, the probability that a randomly chosen set of @ tags matches a randomly chosen set of @ words is simply <Dy .We then have an expression for the chance of match ing a specific alignment of an active state to a specific span.~~
~~Clearly, there can be many alignments which differ only in the spans of the categories, but line up the same tags with the same words.~~
~~However, there will be a certain number of unique ways in which the words and tags can be lined up between ? and = . If we know this number, we can calculate the total probability that there is some alignment which matches.~~
~~For example, consider the state NP NP CC NP . PP (which has signature (1,2) ? the PP has no effect) over a span of length , with empties available.~~
~~The NPs can match any span, so there are alignments which are distinct from the standpoint of the CC tag ? it can be in any position.~~
~~The chance that some alignment will match is therefore OdFPOz<[ I , which, for small < is roughly linear in . It should be clear that for an active state like this, the longer the span, the more likely it is that this state will be found over that span.~~
~~It is unfortunately not the case that all states with the same signature will match a span length with the same probability.~~
~~For example, the state NP NP NP CC . NP has the same signature, but must align the CC with the final element of the span.~~
~~A state like this will not become more likely (in our model) asspan size increases.~~
~~However, with some straightforward but space-consuming recurrences, we can calcu late the expected chance that a random rule of a given signature will match a given span length.~~
~~Since we know how many states have a given signature, we can calculate the total active saturation ?q=Z?&@2G as ?q=?$@2G {flNHN|}o2E uTv@2'p Ffi.~ |?~~
~~?z?&@Fox???[t8G ??~~
~~more efficient than NOUNARIESHIGH despite having moreactive states, largely because using the bottoms of chains in creases the frequency of tags relative to categories.~~
~~16This is also false; tags occur slightly more often at the beginnings of rules and less often at the ends.~~
~~This model has two parameters.~~
~~First, there is < which we estimated directly by looking at the expected matchbetween the distribution of tags in rules and the distri bution of tags in the Treebank text (which is around 1/17.7).~~
~~No factor for POS tag ambiguity was used, another simplification.17 Second, there is the map o2E u[v@ from signatures to a number of active states, which was read directly from the compiled grammars.~~
~~This model predicts the active saturation curves shown to the right in figure 11.~~
~~Note that the model, though not perfect, exhibits the qualitative differencesbetween the settings, both in magnitudes and expo nents.18 In particular: The transform primarily changes the saturation overshort spans, while the encoding determines the over all magnitudes.~~
~~For example, in TRIE-NOEMPTIESthe low-span saturation is lower than in TRIE NOTRANSFORM since short spans in the former case can match only signatures which have both @ and o small, while in the latter only @ needs to be small.~~
~~Therefore, the several hundred states which are reachable only via categories all match every span starting from size 0 for NOTRANSFORM, butare accessed only gradually for NOEMPTIES.~~
~~How ever, for larger spans, the behavior converges to counts characteristic for TRIE encodings.~~
~~For LIST encodings, the early saturations are huge, due to the fact that most of the states which are available early for trie grammars are precisely the ones duplicated up to thousands of times in the listgrammars.~~
~~However, the additive gain over the ini tial states is roughly the same for both, as after a few items are specified, the tries become sparse.~~
~~The actual magnitudes and exponents19 of the sat urations are surprisingly well predicted, suggesting that this model captures the essential behavior.These active saturation curves produce the active total curves in figure 10, which are also qualitatively cor rect in both magnitudes and exponents.~~
~~4.4 Traversals.~~
~~Now that we have models for active and passive edges, we can combine them to model traversal counts as well.~~
~~We assume that the chance for a passive edgeand an active edge to combine into a traversal is a single probability representing how likely an arbitrary active state is to have a continuation with a label match ing an arbitrary passive state.~~
~~List rule states have onlyone continuation, while trie rule states in the branch17In general, the ? we used was lower for not having mod eled tagging ambiguity, but higher for not having modeled the fact that the SCCs are not of size 27.~~
~~18And does so without any ?tweakable?~~
~~parameters.~~
~~19Note that the list curves do not compellingly suggest a power law model.~~
~~0.0M 5.0M 10.0M 15.0M 20.0M 0 10 20 30 40 50 SentenceLength A vg .T ra ve rs al s List-NoTransform exp2.60r0.999 Trie-NoTransform exp2.86r1.000 Trie-NoEmpties exp3.28r1.000 0.0M 5.0M 10.0M 15.0M 20.0M 0 10 20 30 40 50 SentenceLength A vg .T ra ve rs al s List-NoTransform exp2.60r0.999 Trie-NoTransform exp2.92r1.000 Trie-NoEmpties exp3.47r1.000 Figure 13: The average number of traversals for sentences of a given length as observed in practice (left), and as predicted by the models presented in the latter part of the paper (right).~~
~~ing portion of the trie average about 3.7 (min FSAs4.2).20 Making another uniformity assumption, we assume that this combination probability is the contin uation degree divided by the total number of passive labels, categorical or tag (73).~~
~~In figure 13, we give graphs and exponents of thetraversal counts, both observed and predicted, for various settings.~~
~~Our model correctly predicts the approx imate values and qualitative facts, including: For LIST, the observed exponent is lower than forTRIEs, though the total number of traversals is dramatically higher.~~
~~This is because the active satura tion is growing much faster for TRIEs; note that in cases like this the lower-exponent curve will never actually outgrow the higher-exponent curve.~~
~~Of the settings shown, only TRIE-NOEMPTIES exhibits super-cubic traversal totals.~~
~~Despitetheir similar active and passive exponents, TRIE NOEMPTIES and TRIE-NOTRANSFORM vary in traversal growth due to the ?early burst?~~
~~of activeedges which gives TRIE-NOTRANSFORM signifi cantly more edges over short spans than its power law would predict.~~
~~This excess leads to a sizeablequadratic addend in the number of transitions, caus ing the average best-fit exponent to drop without greatly affecting the overall magnitudes.Overall, growth of saturation values in span size in creases best-fit traversal exponents, while early spikes in saturation reduce them.~~
~~The traversal exponents therefore range from LIST-NOTRANSFORM at 2.6 toTRIE-NOUNARIESLOW at over 3.8.~~
~~However, the fi nal performance is more dependent on the magnitudes, which range from LIST-NOTRANSFORM as the worst, despite its exponent, to MIN-NOUNARIESHIGH as thebest.~~
~~The single biggest factor in the time and traver sal performance turned out to be the encoding, which is fortunate because the choice of grammar transform will depend greatly on the application.~~
~~20This is a simplification as well, since the shorter prefixes that tend to have higher continuation degrees are on average also a larger fraction of the active edges.~~
~~We built simple but accurate models on the basis of two observations.~~
~~First, passive saturation is relatively constant in span size, but large due to high reachabilityamong phrasal categories in the grammar.~~
~~Second, ac tive saturation grows with span size because, as spans increase, the tags in a given active edge are more likelyto find a matching arrangement over a span.~~
~~Combin ing these models, we demonstrated that a wide range of empirical qualitative and quantitative behaviors of an exhaustive parser could be derived, including the potential super-cubic traversal growth over sentence lengths of interest.~~