=Paper=
{{Paper
|id=Vol-1240/feosw2014-paper3
|storemode=property
|title=Predicting the impact of central bank communications on financial market investors' interest rate expectations
|pdfUrl=https://ceur-ws.org/Vol-1240/feosw2014-paper3.pdf
|volume=Vol-1240
|dblpUrl=https://dblp.org/rec/conf/esws/MonizJ14
}}
==Predicting the impact of central bank communications on financial market investors' interest rate expectations ==
https://ceur-ws.org/Vol-1240/feosw2014-paper3.pdf
Predicting the impact of central bank communications on
financial market investors’ interest rate expectations
Andy Moniz1 and Franciska de Jong2,3
1
Erasmus Studio, Erasmus University, Rotterdam, The Netherlands
{moniz}@rsm.nl
2
Erasmus Studio, Erasmus University, Rotterdam, The Netherlands
{fdejong}@eshcc.eur.nl
3
Human Media Interaction, University of Twente, Enschede, The Netherlands
{f.m.g.dejong}@utwente.nl
Abstract. In this paper, we design an automated system that predicts the impact
of central bank communications on investors’ interest rate expectations. Our
corpus is the Bank of England’s ‘Monetary Policy Committee Minutes'. Prior
studies suggest that effective communications can mitigate a financial crisis; in-
effective communications may exacerbate one. The system described here
works in four phases. First, the system detects salient aspects associated with
economic growth, prices, interest rates and bank lending using information
from Wikipedia. These economic aspects are detected using the TextRank link
analysis algorithm. A multinomial Naive Bayesian model then classifies docu-
ment sentences to these aspects. The second phase measures sentiment using a
count of terms from the General Inquirer dictionary. The third phase employs
Latent Dirichlet Allocation (LDA) to infer topic clusters that may acts as inten-
sifiers/diminishers for the economic aspects. Finally, an ensemble tree com-
bines the phases to predict the impact of the communications on financial mar-
ket interest rates.
Keywords: sentiment analysis·text mining·link analysis·financial markets
1 Introduction
Post the global financial crisis, there has been a dramatic change in the use of central
bank communications as a central bank policy instrument [1,2]. Central banks com-
municate qualitative information to the financial market through statements, minutes,
speeches, and published reports [3]. Communication is an important tool that a central
bank can use to avert a crisis, by providing investors with its assessment of the risks
and the measures it views as necessary to reduce those risks within the economy [1],
[4]. Previous studies suggest that effective central bank communications can mitigate
and potentially prevent a financial crisis; ineffective communications may exacerbate
one [1], [5]. In [4], the Swedish central bank, the Riksbank, is criticized because its
communications were “not clear or strong enough” leading up to the global financial
crisis, such that the bank’s information went “unnoticed” [1]. In this paper, we design
an automated system that predicts the impact of central bank communications on
interest rate expectations, as derived via financial market patterns. For the purposes of
this study, we analyze economic sentiment, as expressed in the ‘Monetary Policy
Committee Minutes' [2] published by the Bank of England, that details its monthly
interest rate decisions.
Financial markets scrutinize central bank communications for “clues and shades of
meaning about its assessment of the economy and the direction of where economic
policy may be heading” [1]. As a prediction task, the measurement and evaluation of
sentiment is challenging due to the complexities and subtleties of interpreting bank
communications [1]. The formation of economic policy is a balancing act between
achieving high economic growth and financial stability, while targeting low inflation
[2]. The relative importance of these objectives is dynamic, and varies depending on
the prevailing economic conditions [2]. For example under benign economic condi-
tions, high inflation may be construed by financial market investors as a negative
signal for the direction of future interest rates. During the financial crisis of 2007-
2009, high inflation was considered to be a positive signal by effectively lowering
interest rates1 [6]. This motivates a need for fine-grained sentiment analysis, to auto-
matically detect economic aspects and predict the central bank sentiment expressed
towards these aspects [7]. Such a model would provide investors with an automated
system to decipher the complexities and interactions of economic aspects, to interpret
the consequences of these interactions for the future path of interest rates, and to in-
corporate the information into their investment decisions. For a central bank, such a
model would provide it with the ability to predict the impact of its economic policies
on the financial markets. The resulting ‘price discovery’ process [2] may promote a
more efficient functioning of financial markets.
Our approach consists of four phases. First, the system detects salient references to
economic aspects associated with economic growth, prices, interest rates and bank
lending and employs a multinomial Naive Bayesian model to classify sentences with-
in documents. Economic aspects are identified in a pre-processing step, that employs
a link analysis using the TextRank algorithm [8,9]. The second phase measures sen-
timent expressed for the economic aspects, using a count of terms from the General
Inquirer dictionary [17]. The third phase employs Latent Dirichlet Allocation (LDA)
to infer intensifiers/diminishers that may change the meaning of the economic aspects
and economic sentiment [10],[7]. Specifically, the model categorizes whether the
magnitude of the economic aspects has ‘intensified’ or ‘diminished’ over time
[11,12]. We refer to the resulting topic clusters as directional topic clusters. Finally,
an ensemble tree combines the model components to predict the impact of the com-
munications on financial market interest rates over the following day.
1
The real interest rate is the rate of interest a borrower expects to pay on debt after allowing for inflation
and is equal to the nominal interest rate (set by the central bank) minus the rate of inflation [2]
The rest of this paper is structured as follows. Section 2 draws on literature from the
fields of economics and discusses the implications for sentiment analysis and key-
word detection. Section 3 models the individual components of the system. Section 4
outlines the corpus of central bank communications, provides an evaluation of the
model components and then discusses the results. Section 5 concludes and suggests
avenues for future research.
2 Related Work
2.1 Background: central bank research
Post the financial crisis, several central banks have identified communications, partic-
ularly ‘enhanced forward guidance’, as an important policy instrument within their
economic toolkit [1,2]. Effective communications enhance a central bank’s public
transparency, accountability and credibility [13], which in turn aids its ability to im-
plement economic policies [14]. To date, there has been little research into text min-
ing of central bank communications. In [14], the impact of different types of commu-
nications (press releases, speeches, interviews, and news conferences) are analyzed to
determine which media sources impact interest rate expectations. The analysis does
not, however, classify the language used in the documents. In [3], a term counting
approach is adopted to analyze the sentiment contained within the meeting minutes of
the US central bank (the Federal Reserve). In [3] and [15] Latent Semantic Analysis is
employed to analyze the sentiment contained within the Bank of Canada’s minutes.
The intention of this study is to design a fine-grained sentiment analysis approach to
analyze the impact of central bank communications on financial market investors. To
our knowledge, this remains an unexplored avenue of research.
2.2 Background: sentiment analysis
Traditionally, fine-grained sentiment analysis has been researched for the classifica-
tion of online user reviews of products and movies [16]. Readers are often not only
interested in the general sentiment towards an aspect but also a detailed opinion anal-
ysis for each of these aspects [7]. Evaluation is conducted by comparing model classi-
fications versus ratings provided by users. The evaluation of economic sentiment is
arguably a harder task, due to the lack of a clearly defined outcome to assess model
performance. For example, which economic variable should a model’s predictions be
evaluated against? The relative importance of the aspects (e.g. economic growth/
inflation/interest rates) is subjective, may vary over time, and the measurement of the
aspects is only known with significant time delay.
The traditional approach to text-mining within the field of finance is to count terms
using the General Inquirer dictionary [17,18]. The dictionary classifies words accord-
ing to multiple categories, including 1,915 positive words and 2,291 negative words.
The General Inquirer was developed for psychology and sociology research and
while it is used for text mining within the field of finance, little research has been
conducted as to its suitability within finance [19]. Aspects that are frequently men-
tioned in central bank communications, such as the terms ‘employment’, ‘unemploy-
ment’ and ‘growth’, are not classified by the General Inquirer dictionary. Adjectives
are often needed before investors can interpret the patterns in the economy to form
their interest rate expectations [3]. Furthermore, the terms ‘inflation’ and ‘low’ are
classified as negative by the dictionary, yet ‘low inflation’ is a positive characteristic
and indeed achieving this is a central bank’s core objective [2]. The terms ‘fall’ and
‘decline’ are classified as negative terms in the General Inquirer dictionary, yet the
opposite terms ‘rise’ and ‘increase’ are not classified at all.
2.3 Background: keyword detection
Graph-based algorithms have received much attention [8] as an approach to keyphrase
extraction and are considered to be state-of-the-art unsupervised methods [20]. In a
graph representation of a document, nodes are words or phrases, and edges represent
co-occurrence or semantic relations. The underlying assumption is that all words in
the text have some relationship to all other words in the text. Such an approach is
statistical, because it links all co-occurring terms without considering their meaning
or function in text. Centrality is often used to estimate the importance of a word in a
document [22], and is a way of deciding on the importance of a vertex within a graph
that takes into account global information recursively computed from the entire graph,
rather than relying only on local vertex-specific information [23]. The main advantage
of such a representation is that selected terms are independent of their language [21].
3 Model to predict changes in investors’ expectations
In this section we describe the four phases of the system. First, the system detects
salient references to economic aspects and employs a multinomial Naive Bayesian
model to classify sentences within documents. The second phase measures sentiment
expressed for the economic aspects, using a count of terms from the General Inquirer
dictionary. The third phase employs a LDA model and categorizes whether the mag-
nitude of the economic aspects has ‘intensified’ or ‘diminished’ [11,12]. Finally, an
ensemble tree combines the model components to predict the impact of the communi-
cations on financial market interest rates over the following day.
3.1 Aspect detection
In [3] it is shown that tf-idf weighting selects infrequent terms that relate to major
news events or economic shocks. Our approach is intended to detect the common
economic themes that are discussed in central bank communications, that are more
likely to influence investors’ interest rate expectations on a day-to-day basis [2]. To
determine salient references, we employ a link analysis approach that detects the most
frequently mentioned terms within two Wikipedia pages on Central Banking and In-
flation. The model employs TextRank [8], a ranking algorithm based on the concept
of eigenvector centrality, to compute the importance of the nodes in the graph. Each
vertex corresponds to a word. A weight, wij, is assigned to the edge connecting the
two vertices, vi and vj. The goal is to compute the score of each vertex, which reflects
its importance, and use the word types that correspond to the highest scored vertices
to form keywords for the text [23]. The score for vi, S(vi), is initialized with a default
value and is computed in an iterative manner until convergence using recursive for-
mula shown in Equation (1).
(1)
where Adj(vi) denotes vi’s neighbors and d is the damping factor set to 0.85 [8]. Fig-
ure 1 displays the resulting clustering of terms. The size of each node is directly pro-
portional to the TextRank score of the respective economic aspect.
Fig. 1. Link analysis of frequently occurring terms. Different nodes colors reflect different
communities identified using the Clauset-Newman-Moore algorithm.
We define economic aspects by employing a greedy algorithm to detect communi-
ties of terms within the network [31]. The economic growth aspect detects the fre-
quency of the terms: ‘demand’, ‘goods’, ‘services’, ‘investment’. The prices aspect
detects the terms: ‘inflation’, ‘prices’ , ‘money’, ‘markets’, ‘currency’. The interest
rate aspect detects the occurrence of: ‘interest’, ‘rates’, ‘policy’ and a bank lending
aspect detects the terms: ‘banks’, ‘lending’ and ‘assets’. It is not surprising to see
these terms appear in the link analysis, given a central bank’s remit is to maintain
price and financial stability. The choice of terms is consistent with the text mining
research of [3] which identifies 'growth', 'price', 'rate', and 'econom' as the most fre-
quently occurring terms for the US economy. Using the four economic aspects, the
system employs a multinomial Naive Bayesian model [24] to categorize sentences
within each document. The resulting classification labels form the basis upon which
sentiment analysis is applied.
3.2 Polarity detection
In the second phase, the model computes a measure of economic sentiment associated
with each of the economic aspects. We measure polarity by counting the number of
positive (P) versus negative (N) terms, (P − N)/(P + N) identified using the General
Inquirer dictionary [17]. In line with [16], our goal is not to show that a term counting
method can perform as well as a Machine Learning method, but to provide a baseline
methodology to measure central bank sentiment and to draw attention to the limita-
tions of the approach that is widely adopted by text mining studies in the field of fi-
nance as indicated in Section 2.2. The sentiment metrics that are associated with the
economic aspects: economic growth, prices, interest rate and bank lending are la-
belled Tonegrowth, Toneprices, Toneinterest_rates and Tonebank_lending respectively. A fifth
sentiment metric, Toneoverall, is computed to measure the polarity associated with the
overall document, without conditioning upon the economic aspects. The five senti-
ment metrics are included as separate components within the ensemble tree.
3.3 Detection of LDA directional topic clusters
Next we extend the baseline term-counting method by taking intensifiers and dimin-
ishers into account [11,12]. These are terms that change the degree of the expressed
sentiment in a document (see Section 2.2). In the case of central bank communica-
tions, the terms describe how economic aspects have changed over time. We employ
an implementation of LDA [10], and represent each document as a probability distri-
bution over latent topics, where each topic is modeled by a probability distribution of
words. In [7], LDA is found to capture the global topics in documents, to the extent
that topics do not represent ratable aspects associated with individual documents, but
define clusterings of the documents into specific types. For the purposes of training
the LDA model, we consider each sentence within each central bank communication
to be a separate document. This increases the sample size of the dataset (see Section
4.1) and is intended to improve the robustness of the LDA model for statistical infer-
ence. We implement standard settings for LDA hyper-parameters, α = 50/K and
β=.01, where the number of topics K is set to 20 [25]. We manually annotate two of
the topic clusters that capture ‘directional’ information [1] and appear to act as inten-
sifiers/diminishers of meaning. We label the clusters directional topic clusters. Table
1 identifies the top terms associated with the two clusters. Representative words are
the highest probability document terms for each topic cluster.
Table 1. Representative document terms associated with the directional topic clusters
'intensifier cluster' 'diminisher cluster'
word prob. word prob.
increase 0.150 moderated 0.190
strong 0.107 lower 0.161
accelerate 0.081 downwards 0.123
strength 0.063 difficult 0.102
support 0.058 less 0.070
Next for each central bank communication the LDA model infers the probabilities
associated with the ‘intensifier’ and ‘diminisher’ clusters within each of the three
economic aspects detected by the Naïve Bayesian classifier. The output of the model
is a vector of six topic probabilities that proxy the central bank’s assessment that the
economic aspects are intensifying/diminishing. We label the model directional LDA
model and the respective probability vectors: Topicgrowth, Topicprices, Topicinter-
est_rates and Topicbank_lending if the economic aspects are increasing and Topicgrowth,
Topicprices, Topicinterest_rates and Topicbank_lending if the economic aspects are decreas-
ing. We include the topic probabilities as components within the ensemble tree.
4 Experiments
In this section we discuss the corpus of central bank communications and describe the
investor patterns data used to evaluate the impact of the central bank communications
on investors’ interest rate expectations. We then outline the evaluation of the ensem-
ble classification tree, present the results and provide a discussion.
4.1 Data
We choose to analyze the interest rate minutes of the Bank of England. As cited in
[3], central bank minutes are closely watched by investors to gauge the future direc-
tion of economic policies. Similar datasets for the US and Canadian central banks’
minutes are examined in [3] and [15]. The Bank of England announces the level of
UK interest rates on the first Thursday of every month. The details that underpin this
decision are only provided two weeks later and are published in the Bank of Eng-
land’s ‘Monetary Policy Committee Minutes’. The communications are interesting to
analyze, because changes in investors’ expectations on the day of the central bank
communication may be attributed to the qualitative information contained within the
meeting minutes rather than the interest rate decision announced two weeks before.
Minutes typically include summaries of committee members’ views on economic
conditions and discuss the rationale for their interest rate decisions [26]. The central
bank’s minutes are, on average, 12 pages long (including a header page), and contain
around 55 bullet points, typically with 5 sentences in each bullet. The documents are
available from 1997, the year when Parliament voted to give the Bank of England
operational independence from the UK government. We retrieve all meeting minutes
available between July 1997-March 20142 to create a corpus that consists of 199 doc-
uments. For the purposes of aspect detection and to train the LDA model, we remove
the header page and define a document as an individual sentence within each of the
meeting minutes. This expands the corpus to a collection of 53,195 documents.
To evaluate the ensemble tree’s predictions we utilize information obtained from
financial market patterns. Interest rate futures contracts are financial instruments that
enable investors to insure against or speculate on uncertainty about the future level of
interest rates [27]. Changes in the price of the futures contracts therefore reflect
changes in investors’ views on the future direction in central bank interest rates. In-
vestors’ interest rate expectations for the following three, six and twelve months are
derived and published daily by the Bank of England. We utilize investors’ twelve
month ahead forecasts. This data series has the greatest data coverage compared to the
three and six month series. Furthermore, the twelve month forecast horizon is con-
sistent with the time horizon over which that the Bank of England conducts its eco-
nomic policies [2]. To isolate the effect of the central bank communication on inves-
tors’ expectations, we compute the percentage change in the interest rate futures con-
tract, as measured from the close of business on the day of the communication an-
nouncement until the close of business one day after. This narrow time window helps
to minimize the influence on investors’ interest rate expectations from other financial
market factors that may occur at the same time [28].
4.2 Experiment setup
We design the evaluation of the prediction model in stages to enhance our understand-
ing of the model components. For a baseline, we evaluate the system’s predictions by
using only the tone of the overall document (see Section 3.2). The approach does not
take into account individual economic aspects or diminishers/intensifiers [11,12]. We
label the model naïve tone. This approach is consistent with the methodology adopted
by the financial literature [18]. Next we compare the outcomes of an ensemble model
that combines the tone associated with each of the economic aspects: economic
growth, inflation and interest rates (see Section 3.2). We label this the economic as-
pects model. A third model compares the outcomes from an ensemble model that
combines the tone of eight directional economic aspects, that combines the intensifi-
ers/diminishers associated with the four economic aspects (see Section 3.3). We label
this the directional LDA model. Finally, we combine the components in a single en-
semble tree and refer to the system as the joint aspect-polarity model.
Learning and prediction is performed using an ensemble tree. The goal of ensemble
methods is to combine the predictions of several models built with a given learning
algorithm in order to improve generalizability and robustness over a single model. We
use the Random Forest algorithm [30], that employs a diverse set of classifiers by
introducing randomness into the classifier construction. Experiments were validated
2
Central bank communications announced in August 1997 were excluded from the analysis because the
communication document was not readily available in a machine readable format.
using five-fold cross validation in which the dataset is broken into five equal sized
sets; the classifier is trained on four datasets and tested on the remaining dataset. The
process is repeated five times and we calculate the average across folds. For evalua-
tion, we select Mean Absolute Error (MAE), Root Mean Squared Error and Spear-
man’s rho (ρ). We also examine Spearman's rho since prediction may be considered
to be a ranking task. The formulae are displayed in Equation (2) below.
(2)
where Ei is the model’s predicted value, Oi is the realized value, and n is the number
of observations. MAE measures the average magnitude of the forecast errors without
considering direction; RMSE penalizes errors and gives a relatively high weight to
large errors. A smaller value of MAE or RMSE indicates a more accurate prediction.
Spearman's rho is a non-parametric measure of the degree of linear association be-
tween the predicted and realized values, and is bound between the range -1 to +1 [29].
A positive Spearman's rho indicates the model’s predictive ability; a negative value
indicates a poor model fit.
4.3 Experiment results
The evaluation metrics from the model components are shown in Table 2.
Table 2. Evaluation of the model components
Model MAE RMSE ρ
***
naive tone 0.022 0.016 -0.187
economic aspects 0.018 0.013 -0.044
directional LDA 0.019 0.014 0.041
joint polarity model 0.015 0.011 0.034
The asterisks provide the level of significance where *** indicates that the model’s predictions
versus forecasts are statistically, negatively significant at the 0.1% level.
The naïve tone model, which is similar to the approach commonly adopted by text
mining studies in the field of finance, shows the worst performance. It exhibits the
highest MAE and RMSE. The rank correlation of the model’s forecasts with realized
changes in investors’ interest rate expectations is negative and is highly statistically
negative, implying that documents that are predicted to have a positive/negative im-
pact on investors’ interest rate expectations end up having the reverse effect. The
economic aspects and directional LDA models exhibit monotonic decreases in MAE
and RMSE, suggesting a slight improvement in the model fit. Spearman’s rho, how-
ever, is again negative, albeit to a lesser extent. Finally, the joint aspect-polarity mod-
el, that includes all model components in the ensemble tree, displays the lowest MAE
and RMSE. The mildly positive Spearman’s rho is consistent with previous studies
within the field of finance. As cited in [19], many factors influence the financial mar-
kets; a low, positive correlation provides sufficient comfort of the model’s predictive
power.
4.4 Discussion
One interpretation of the experiment results is that multiple aspects are needed to
improve the accuracy of the system. The positive Spearman’s rho for the joint model
versus the negative Spearman’s rho for the naïve tone, economic aspects and direc-
tional LDA models may be indicative of a non-linear relationship between the com-
ponents that is only evident when the models are combined rather than considered in
isolation. One of the strengths of a regression tree is that it does not assume a func-
tional form, allowing it to detect interactions between model components. To aid our
understanding of prediction in the joint model, Figure 2 displays the decision tree
results for one of the folds. The values in the grey boxes provide the predicted per-
centage change in investors’ interest rate expectations associated with the sentiment
contained within the central bank communication. A positive value indicates that the
impact is expected to lead to an increase in investors’ interest rate expectations, while
a negative value indicates an expected decrease in interest rate expectations.
Fig. 2. Example decision tree from one of the folds
The regression tree identifies the interaction between the directional topic clusters
and Tone measures. The primary decision in the decision tree is central bank senti-
ment towards economic growth. The right hand path indicates that if a central bank
communication emphasizes positive economic growth and discusses interest rate in-
creases, the model predicts that investors’ expectations of future interest rates will
rise by 3%. The left hand path predicts that if a central bank tone towards economic
growth is low, declining bank lending and the tone towards interest rates is negative,
investors will reduce their expectations of future interest rates by 4%.
5 Conclusion
The goal of central bank communication is to make messages as clear, simple and
understandable as possible to a wide range of audiences [1]. In this study, we focus of
one specific audience, namely financial market investors. Investors play a key role for
the implementation of a central bank’s economic policies [1,2]. The outcome of our
study may feed the design of a system that can predict the impact of central bank
communication on formation of investors’ interest rate expectations. The results of
the joint aspect-polarity model suggest that investors may benefit by incorporating a
measure of central bank sentiment to forecast interest rates.
In this study we evaluate model performance using prices from financial market
instruments. The market price of an interest rate contract implicitly measures the
average investor's interest rate expectations [27]. It is also possible to compute an
'implied probability distribution' of those expectations [27]. In future work we plan to
evaluate a range of metrics, including the dispersion of the expectations as a proxy of
investor uncertainty. Post the 2007–09 financial crisis, central banks have broadened
the range of their communication, including the use of social media, live broadcasts,
podcasts and blogs, to deliver their messages quickly and efficiently [1]. In future
research, a wider range of central bank communications, including those expressed
via social media, will be integrated into our study. We also intend to examine alterna-
tive approaches to select economic aspects, including dynamic approaches that reflect
the usage of terms as central bank communications change over time.
Acknowledgement
The research leading to these results has partially been supported by the Dutch na-
tional program COMMIT.
References
1. Vayid, I. Central Bank Communications Before, During and After the Crisis: From Open-
Market Operations to Open-Mouth Policy, Bank of Canada Working Paper (2013)
2. Bank of England. Monetary policy trade-offs and forward guidance (2013)
3. Boukus, E., Rosenberg, J., V. The information content of FOMC minutes (2006)
4. Meyersson, P., Karlberg, P.P, A Journey in Communication: the Case of the Sveriges
Riksbank SNS Förlag (2012)
5. Viñals, J. Lessons from the Crisis for Central Banks. IMF Speech (2010)
6. Danthine, J, P. Causes and consequences of low interest rates. Speech by Mr Jean-Pierre
Danthine, Vice Chairman of the Governing Board of the Swiss National Bank, at the
Swisscanto Market Outlook 2014, Lausanne ( 2013)
7. Titov, I, McDonald, R., T. Modeling online reviews with multi-grain topic models. Pro-
ceeding of the 17th WWW (2008)
8. Mihalcea, R., and Tarau, P. TextRank: Bringing order into texts. In Proceedings of the
2004 Conference on Empirical Methods in Natural Language Processing (2004)
9. Brin, S., Page, L. The anatomy of a large-scale hypertextual web search engine. Comput.
Networks ISDN Systems, 33:107–117 (1998)
10. Blei, D., M., Ng, A., Jordan, M., I. Latent Dirichlet Allocation, Journal of Machine Learn-
ing Research 3, 993-1022 (2003)
11. Kennedy, A., Inkpen, D. Sentiment Classification of Movie Reviews using Contextual Va-
lence Shifters. Computational Intelligence, vol.22(2), pp.110-125 (2006)
12. Polanyi, L., Zaenen, A. Contextual Valence Shifters, AAAI Spring Symposium on Explor-
ing Attitude and Affect in Text: Theories and Applications (2004)
13. Carney, M. Panel discussion comments to the BIS Conference on The Future of Central
Banking under Post-Crisis Mandates (2010)
14. Fay, C., Gravelle, T. “Has the Inclusion of Forward-Looking Statements in Monetary Poli-
cy Communications Made the Bank of Canada More Transparent?” Bank of Canada Dis-
cussion Paper (2010)
15. Hendry, S., Madeley, A. Text Mining and the Information Content of Bank of Canada
Communications, Bank of Canada Working Paper (2010)
16. Pang, B., Lee, L., Vaithyanathan, S. Thumbs up? Sentiment classification using machine
learning techniques. Proceedings of EMNLP-02 (2002)
17. Stone, P., Dumphy, D. C., Smith, M. S., and Ogilvie, D. M. The General Inquirer: A Com-
puter Approach to Content Analysis. The MIT Press. (1966)
18. Tetlock, P., Saar-Tsechansky, M. Macskassy, S. More Than Words: Quantifying Language
to Measure Firms’ Fundamentals. Journal of Finance, Vol. LXIII, No. 3, 1437-1467 (2008)
19. Loughran, T., McDonald, B. When is a liability not a liability? Textual analysis, dictionar-
ies and 10Ks. Journal of Finance 66, 35–65 (2010)
20. Liu, F. , Pennell, D., Liu, F., Liu, Y. Unsupervised approaches for automatic keyword ex-
traction using meeting transcripts. In Proceedings of Human Language Technologies: The
2009 Annual Conference of the North American (2009)
21. Litvak, M., Last, M. Graph-Based Keyword Extraction for Single-Document Summariza-
tion. Proceedings of the 2nd Workshop on Multi-source, Multilingual Information Extrac-
tion and Summarization, Coling 2008, pp. 17-24. Association for Computational Linguis-
tics (2008)
22. Opsahl, T., Agneessens, F., Skvoretz, J. Node centrality in weighted networks: Generaliz-
ing degree and shortest paths (2010)
23. Boudin, F. A Comparison of Centrality Measures for Graph-Based Keyphrase Extraction
(2013)
24. McCallum, A., and Nigam, K. A comparison of event models for naive Bayes text classifi-
cation. AAAI-98 Workshop on Learning for Text Categorization (1998)
25. Griffiths, T. L., Steyvers, M. Finding scientific topics. Proceedings of the National Acad-
emy of Science, 101, 5228-5235 (2004)
26. Danker, D., J., and Luecke, M., M. Background on FOMC Meeting Minutes, Federal Re-
serve Bulletin 175-179 (2005)
27. Clews, R., N. Panigirtzoglou, Proudman, J. Recent developments in extracting infor-
mation from options markets, Bank of England Quarterly Bulletin (2000)
28. Mackinlay, A., C. Event Studies in Economics and Finance. Journal of Economic Litera-
ture (1997)
29. Maritz, J., S. Distribution-free statistical methods. Chapman and Hall, (1984)
30. Breiman, L. Random Forests. Machine Learning (2001)
31. Clauset, A., Newman, M. E. J., and Moore, C. Finding community structure in very large
networks. Physical Review (2004)