Using news to predict Chinese medicinal material price index movements

Publication Date11 Jun 2018
AuthorMiao Yu,Chonghui Guo
SubjectInformation & knowledge management,Information systems,Data management systems,Knowledge management,Knowledge sharing,Management science & operations,Supply chain management,Supply chain information systems,Logistics,Quality management/systems
Using news to predict Chinese
medicinal material price
index movements
Miao Yu and Chonghui Guo
Institute of Systems Engineering, Dalian University of Technology, Dalian, China
Purpose The purpose of this paper is to propose an approach for predicting the movements of Chinese
medicinal material price indexes using news based on text mining.
Design/methodology/approach A research framework and three major methods, namely, domain
dictionary construction, market convergence time calculation and dimensionality reduction integrating
semantic analysis,are proposed for the approach. The proposedapproach is applied in practice for predicting
the price indexmovements of the top ten Chinesemedicinal materials that receivethe greatest media attention.
Findings A set of experimentsperformed hereinshow that a predictive relationshipexists between the news
and the commodity market and that each of the threemajor methods improves the forecasting performance.
Research limitations/implications Because the field of Chinese medicinal materials lacks a corpus that
can be used for sentiment analysis, the accuracy of a trained automatic sentiment classifier is lower than
obtained by a manual method, which can cause the calculated convergence result to be inaccurate, thus
affecting the final prediction model. The manual method of having people label news decreases the proposed
methods aspects of being intelligent and automatic.
Practical implications Using the method proposed herein to predict the trends in Chinese medicinal
materials prices helps farmers arrange a reasonable planting plan to pursue their best interests.
Social implications The method proposed herein to predict the trends in the prices of Chinese medicinal
materials is conducive to the government arranging planned drug availabilities in order to avoid disasters in
which herbs are looted.
Originality/value The produced prediction result is meaningful in supporting farmers and investors to
make better decisions in growing and trading Chinese medicinal material, which leads to financial returns on
investments and the avoidance of severe losses.
Keywords Text mining, Chinese medicinal material price index, Movement prediction
Paper type Research paper
1. Introduction
In recent years, the prices of Chinese medicinal materials have fluctuated violently and
frequently, which has had a negative impact on farmers and investors. The ability to
forecast future price changes can support farmers and investors in making better decisions
about growing and trading Chinese medicinal material, which leads to financial returns on
investments and the avoidance of severe losses. Therefore, it is of great interest to predict
Chinese medicinal material price movements.
Due to fluctuations in the prices of Chinese medicinal materials as a result of natural
disasters, acute infectious disease outbreaks and other severe emergencies, the simple
use of historical data cannot be used to predict natural disasters or acute infectious
diseases such as in outbreaks; as a result, we cannot accurately predict the price
fluctuations of Chinese medicinal materials. While, when natural disasters or acute
infectious diseases and other emergencies occur, these will be related to the timely release
of news reports, and with the use of text mining technology to analyse the relevant
Industrial Management & Data
Vol. 118 No. 5, 2018
pp. 998-1017
© Emerald PublishingLimited
DOI 10.1108/IMDS-06-2017-0287
Received 30 June 2017
Revised 9 November 2017
22 December 2017
Accepted 31 December 2017
The current issue and full text archive of this journal is available on Emerald Insight at:
The abstract version was presented at the 1st International Symposium of Supply Chain and Service
Innovation, Guangzhou, China, 67 April, 2017. This research was supported by the National Natural
Science Foundation of China (Grant Nos 71771034 and 71421001).
This paper forms part of a special section Featured issue on supply chain innovation.
real-time news, we can mine the trends in Chinese medicinal materials prices to predict
the directions of its fluctuations. In the field of stocks and foreign exchange, text mining
technology has been to predict the relevant information and has achieved effective
predictive results.
When news is used to predict the market by text mining, the market convergence time
must be determined first. The market convergence time is the shortest time required for
the price to adequately reflect the information available to market participants. After
knowing the market convergence time, the news that really influences the price will be
obtained. Then, the information can be processed by text mining to identify a statistical
relationship between the news and the market movements. Although text mining for
market prediction has received increasing attention by researchers, most of their studies
are about capital markets (stock) or foreign exchange market predictions in multiple types
of financial markets; no one has explored Chinese medicinal material market commodity
markets. Researchers have reported the market convergence time to be 60 minutes
(Chordia et al., 2005). However, in various types of markets, the market convergence time
is expected to be different. Thus, we propose a method for calculating the convergence
time applied to forecasting the commodity market by text mining. In addition,
we addressed the two challenges of retrieving news used for the prediction model more
completely and integrating semantic analysis to reduce the number of features more
effectivity when addressing Chinese text.
The remainder of this paper is organised as follows. In the next section, we summarise
the key related works. Section 3 presents the general flow of text mining for market
prediction. We provide a framework for using news to predict Chinese medicinal materials
price index movements and the proposed methods are described in detail in Section 4.
We illustrate the performance of the proposed model using a case study about the Chinese
medicinal material market in Section 5. The conclusions and future research are discussed
in Section 6.
2. Literature review
We summarise the key works are relevant for two aspects: Chinese medicinal materials
prediction and text mining for market prediction. Table I provides details in those papers for
a thorough review and taxonomy.
2.1 Chinese medicinal materials prediction
Mao and Chang (2014) collected Kang Mei Chinese Medicinal Material monthly data from
September 2012 to June 2013 to predict the Chinese medicinal material price index.
The authors established the grey model (1, 1) (GM) by a difference equation in a discrete
form as a prediction model. Wang et al. (2016) forecasted the monthly price of
Notoginseng Radix et Rhizoma using an autoregressive integrated moving average
model (2, 1, 3) (ARIMA) based on price data from February 2004 to August 2015.
However, these studies ignored the important fact that the media has a strong effect on
the behaviour of the investor, which causes the prices to move (Yang, 2007; You and Wu,
2012; Liu et al., 2014). Investors make investment decisions based on their perception of
the world, while their perception is limited to the information that is available to market
participants. Information constantly is made available through news releases.
When there are favourable or unfavourable news releases, investors make buy or sell
trading choices, respectively. There is no doubt that news has an impact on
investor expectations regarding future price changes and that the trading of these
market participants moves prices to a new equilibrium (Nofsinger, 2001;
Chi and Zhuang, 2009).
material price

To continue reading

Request your trial

VLEX uses login cookies to provide you with a better browsing experience. If you click on 'Accept' or continue browsing this site we consider that you accept our cookie policy. ACCEPT